WO2009052559A1 - A diagnostic assay - Google Patents

A diagnostic assay Download PDF

Info

Publication number
WO2009052559A1
WO2009052559A1 PCT/AU2008/001556 AU2008001556W WO2009052559A1 WO 2009052559 A1 WO2009052559 A1 WO 2009052559A1 AU 2008001556 W AU2008001556 W AU 2008001556W WO 2009052559 A1 WO2009052559 A1 WO 2009052559A1
Authority
WO
WIPO (PCT)
Prior art keywords
snps
medicament
subject
full
mutations
Prior art date
Application number
PCT/AU2008/001556
Other languages
French (fr)
Inventor
Slave Petrovski
Cassandra Esther Irene Szoeke
Leslie Jon Sheffield
Terence John O'brien
Original Assignee
Melbourne Health
Murdoch Childrens Research Institute
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from AU2007905783A external-priority patent/AU2007905783A0/en
Application filed by Melbourne Health, Murdoch Childrens Research Institute filed Critical Melbourne Health
Publication of WO2009052559A1 publication Critical patent/WO2009052559A1/en

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
    • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
    • C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
    • C12Q1/6883—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/106—Pharmacogenomics, i.e. genetic variability in individual responses to drugs and drug metabolism
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/156—Polymorphic or mutational markers
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/158—Expression markers

Definitions

  • the present invention relates generally to diagnostic assays, therapeutic protocols and medicinal predictor model validation. More particularly, genetic tests are provided which determine the suitability of a medicament in the treatment or prophylaxis of a neurological condition. Even more particularly, the present invention provides genetic assays to measure the potential for, or likelihood of, a subject having a positive or adverse treatment response to a neurological medicament. The present invention is particularly useful in the practice of personalized medicine.
  • Epilepsy is estimated to affect approximately 50 million people globally according to statistics provided by the World Health Organization (WHO; 1995). Current epilepsy treatment is less than satisfactory, with 40% of patients having a significant adverse drug reaction (ADR) and 20-40% experiencing seizure recurrence (Mattson et al, The New England Journal of Medicine 575:145-151, 1985; Kwan and Brodie, The New England Journal of Medicine 342:314-319, 2000).
  • AED anti-epileptic drug
  • intra-individual variability in response between different drugs.
  • AED anti-epileptic drug
  • Uncontrolled seizures cost US$2,250 to $3,205 per year per patient (Begley and Beghi, Epilepsia 41:342-219, 2002). Uncontrolled seizures also have includes indirect costs such as loss of productivity and opportunity which account for 70-85% of total disease-related costs (Akobundu et al, PharmacoEconomics 24:869-890, 2006).
  • ADRs and pharmacoresistance to AED medication are affected by the action of multiple genes, including those involved in the metabolic pathway, mode of adsorption, transportation and receptors of the AEDs, as well as immunological processes.
  • the principal behind pharmacogenetics is that mutations within these genes could lead to their malfunction by interfering with their effect on, or of, the AED resulting in an altered function (pharmacoresistance or level of pharmacosensitivity) or lead to increased blood levels of these drugs or enhanced unwanted pharmacodynamic or immunological responses, predisposing the individuals to a greater risk of developing ADRs.
  • candidate genes which mediate the aforementioned outcomes are identified from the pharmacokinetics and pharmacodynamics of AEDs and genes involved in the pathophysiology of epilepsies.
  • the present invention employs a pharmacogenomic approach to screen datasets of two or more genetic markers, such as genes, and/or two or more single nucleotide polymorphisms (SNPs) in genes as a predictor of the likelihood or otherwise that a subject will favorably respond to a drug used to treat a neurological condition.
  • Pharmacoresistance or pharmacosensitivity to a particular medicament used for a neurological disease is proposed herein to be predictable based on the presence or absence of a dataset of SNPs in selected genes. Even more particularly, the neurological disease is epilepsy or a related condition.
  • the present invention provides a pharmacogenomics approach in the assessment of therapeutic outcome potential or pharmacoresistance or pharmacosensitivity to a particular neurological drug.
  • Reference to "pharmacogenomics" in this context includes the study of a spectrum of genes which potentially influences a drug response in a subject.
  • the present invention employs a multivariate approach to pharmacoresistance determination and assessment.
  • the multivariate approach is a predictor of a combinatory genetic effect of drug response.
  • the present invention provides a method for generating a validated medicinal predictor model for use in personalized medicine and epidemiological studies of population groups.
  • the diagnostic assay of the present invention enables a practitioner to select a particular drug or avoid a drug to ensure a reduced likelihood of development of an adverse drug reaction (ADR) in a subject. This leads to greater opportunity for successful treatment, improved quality of life to the subject and reduced health costs. It also enables the practitioner to predict, with an increased degree of certainty, the chance that a patient will have recurrent seizures or other neurological symptoms despite the commencement of drug treatment. This has important implications in providing advice to a patient in relation to daily activities.
  • ADR adverse drug reaction
  • one aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
  • a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
  • reference to a "positive” or “adverse” treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
  • Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
  • Neurological condition includes a neurological, psychiatric or psychological condition, phenotype or state including a sub-threshold neurological, psychiatric or psychological condition, phenotype or state.
  • the neurological condition is epilepsy or a related condition.
  • Reference to a “mutation” generally includes a nucleotide polymorphism such as a single nucleotide polymorphism (SNP) and a multi-nucleotide polymorphism (MNP).
  • SNP single nucleotide polymorphism
  • MNP multi-nucleotide polymorphism
  • Tables 2 A through 21 Datasets of genes or mutations in genes are set forth in Tables 2 A through 21. Reference to "Table 2" includes any or all of Tables 2A through 21. Table 2A includes Table 2A 1 and Table 2A 11 . These lists of genes and mutations represent a first knowledge base. The correlation or weighting between selected mutations in selected genes to a pharmacogenomic effect on medication represents a second knowledge base. The correlation may be determined via an algorithm which represents a training tool.
  • the present invention contemplates use of a set of genes in a first knowledge base to correlate mutations in the form of a second knowledge base via a training tool (i.e. hybrid classifier system) with pharmacosensitivity, pharmacoresistance or ADR to a drug used in neurological treatment.
  • a training tool i.e. hybrid classifier system
  • a subset of genes or mutations in genes is selected based on the drug employed. It is proposed herein that combinations of nucleotide mutations such as SNPs is collectively more highly predictive of a responder compared to univariate analysis based on a single mutation.
  • one aspect of the present invention provides a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, said method comprising screening for the presence of two or more mutations in genes selected from GABBR2, KCNQl, SCN4B and SLC1A3 which mutations correlate to potential responsiveness of the subject to the medicament.
  • mutations in one or more of KCNCl /MYODl, GRIA4 and/or GSTA4 are screened in addition to one or more of GABBR2, KCNQl, SCN4B and/or SLCl A3.
  • the SNPs contemplated herein comprise two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3).
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLCl A3 rs4869682
  • the SNPs contemplated herein are one or more of rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) in combination with one or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3).
  • the SNPs contemplated herein are two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4).
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLCl A3
  • rs3911833 rs3911833
  • KCNCl /MYODl rs507450
  • GRIA4 rsl 82623
  • one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
  • Reference to "two or more” in relation to the latter aspects of the invention includes 2, 3, 4, 5, 6, 7 or 8 SNPs. Reference to one or more includes 1, 2, 3, 4, 5, 6, 7 or 8 SNPs.
  • the second knowledge base may be further particularized into an optimized training set.
  • Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs in genes selected from the list comprising rs2808526
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • rs658624 and rs678262 both in SCN4B
  • rs4869682 SLC1A3
  • rs3911833 KNCl /MYODl
  • rs507450 GRIA4
  • rsl 82623 GTA4
  • rs2808526 GBBR2 also known as GPR51
  • rs2283170 KCNQl
  • rs658624 and rs678262 both in SCN4B
  • rs4869682 SLC1A3
  • rs3911833 KCNCl /MYODl
  • rs507450 GRIA4
  • GSTA4 rs2808526
  • a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • SNPs correlate to potential responsiveness of the subject to the medicament.
  • reference to a "positive" or "adverse” treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
  • Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs
  • Kits, computer programs and treatment and diagnostic protocols are also encompassed by the present invention.
  • a treatment protocol comprises detecting or predicting that a subject will likely be a responder to a particular drug or panel of drugs and then selecting the appropriate drug.
  • a "computer program" includes web-based assays where SNP information is provided to a web site which determines the likelihood of pharmacosensitivity, pharmacoresistance or an ADR.
  • a business method comprising inputting into a web-based site information concerning the presence of two or more mutations in an optimized training set of genes selected from Table 2 wherein the web-based site provides an interactive response providing information on potential pharmacosensitivity or pharmacoresistance or an ADR to a drug used in neurological treatment.
  • mutations in genes includes two or more ofrs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
  • FDA anti-epileptic drug
  • AED anti-epileptic drug
  • CACNA2D1 NM 000722 intronless 7 - chr7:81223068-81727682
  • EPIM NM 194356 func only 12 - chrl2:129798027-129858691
  • GRM8 NM 000845 func only 7 - chr7: 125671607-126487261 Gene name ⁇ a) ref seq (b) gene status (c) Chr (d) Strand (e) Genomic address* 0
  • HTR5A NM 024012 func only 7 + chr7:154290193-154316107
  • NTRK2 NM 006180 func only 9 + chr9:84503019-84869059
  • TRAPPC4 NM_016146 func only 11 + chrl 1 :118384450-1 18400592
  • VAMP2 NM 014232 func only 17 chrl7:8002188-8017017
  • VTIlA NM 145206 func only 10 + chrl ⁇ : l 14187005-114488522
  • Refseq RefSeq Genes refers to known protein-coding genes taken from NCBI mRNA reference sequences collection.
  • Gene status Full: gene considered for both functional and tagger SNP selection.
  • Genomic address The region considered for either or both of functional and tSNP selection.
  • a polymorphism includes a single nucleotide polymorphism (SNP) as well as two or more polymorphisms
  • reference to “an adverse drug reaction” or “an ADR” includes a single ADR, as well as two or more ADRs
  • reference to “the invention” includes a single aspect or multiple aspects of an invention; and so forth.
  • the present invention provides datasets of target genes (first knowledge base) or mutations in target genes wherein two or more mutations enable a correlation (second knowledge base) to be made with respect to the pharmacoresistance or pharmacosensitivity potential of a neurological medicament via an algorithmic training tool referred to as the hybrid classifier system.
  • Candidate genes and/or SNPs are provided in Tables 2A (including Table 2A 1 and 2A 11 ) through 21 which represent training sets of data.
  • a mutation is screened for in two or more of GABBR2, KCNQl, SCN4B and SLCl A3.
  • a mutation in one or more of KCNCl /MYODl, GRIA4 and GST A4 may be screened for in combination with a mutation in one or more of GABBR2, KCNQl, SCN4B and SLCIA3.
  • Optimize training sets include two or more of the SNPs rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • the present invention correlates genotype with predicted treatment outcomes.
  • the correlation is via a bioinformatic analysis of genetic and clinical data in a pharmacogenomic approach to personalized medicine. It is proposed that individual genetic variation has multi-factorial implications in relation to drug absorption, distribution, metabolism, efflux, elimination and variability of drug target receptors which collectively or individually influence treatment outcomes.
  • one or more of the above mutations may be detected in combination with a mutation in MDRl such as the SNP rslO45642.
  • a hybrid univariable/multivariate classification system is contemplated herein predictive of neurological disease treatment outcomes. Prediction rates of 70% or greater are provided herein using the hybrid classifier system. Clinicians can use this system to design personalized treatment programs for individual or cohort patients with reduced incidences of adverse drug reactions (ADRs) and/or poor responders. This increases the overall likelihood of treatment or prophylaxis success, improves health quality of the patient and reduces health costs.
  • ADRs adverse drug reactions
  • Reference to 70% or greater includes 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 and 100%.
  • the present invention further contemplates profiling or stratifying an individual or group of individuals with respect to therapeutic outcome potential in response to a particular drug or class of drugs in the treatment or prophylaxis of a neurological condition. Genotyping with respect to nucleotide mutations creates a genetic profile of a subject and this correlates to a likelihood of the subject responding favorably or not responding (e.g. having recurrence of symptoms or having an ADR) to a particular medicament.
  • a “genetic profile” is meant that an individual or groups of individuals exhibiting a particular neurological condition which includes a neurological, psychiatric or psychological condition, phenotype or state or sub-threshold forms thereof or who are at the risk of developing same, exhibit two or more mutations at or within one or more genes selected from the list in Table 2 including its 5' or 3' terminal regions, promoter, exons or introns which is predictive of a therapeutic outcome.
  • the genetic profile may be a single polymorphism (SNP) or mutli-nucleotide polymorphisms (MNPs) in a single gene or in a panel of genes, that is statistically significantly linked to a neurological condition.
  • a mutation in this context includes a mutation.
  • a mutation also includes a nucleotide insertion, addition, substitution and deletion as well as a rearrangement or microsatellite.
  • Particular genes and/or mutations are provided in Tables 2A through 21.
  • the SNPs contemplated herein comprise two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • Optimized training sets of genes or mutation may differ with a particular population or sub-population including a geographical or ethnic sub-group of a population.
  • One particular optimized training set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLC1A3
  • rs3911833 KNC1/MY0D1
  • rs507450 GRIA4
  • GSTA4 rsl82623
  • one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
  • the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2 which mutations correlate to potential responsiveness of the subject to the medicament.
  • a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
  • reference to a "positive” or “adverse” treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
  • Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2 which mutations correlate to potential responsiveness of the subject to the medicament.
  • the present invention further contemplates a method for identifying a genetic profile in a subject or group of subjects associated with the likelihood of a successful therapeutic outcome or otherwise to a neurological condition, the method comprising screening individuals for two or more polymorphisms including a mutation in a gene selected from the list in Table 2, including its 5' and 3' terminal regions, promoter, introns and exons which has a statistically significant linkage or association to a therapeutic outcome.
  • Table 2 includes Tables 2A through 21 which represent training set of data.
  • Table 2A includes Tables 2A 1 and 2A 11 . It also includes particular SNPs such as optimized training sets selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (G
  • GBR2
  • the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLCl A3 rs4869682
  • a further aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLCl A3
  • rs3911833 rs39
  • Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLCl A3 rs4869682
  • a further aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Still a another aspect of the present invention provides a method for a genotype- based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLCl A3
  • rs3911833 r
  • a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLC1A3 rs4869682
  • Another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Still a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs658624 and rs67826
  • SLC1A3 rs3911833
  • KCNCl /MYODl rs507450
  • GRIA4 rsl826
  • one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
  • reference to a "positive” or “adverse” treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
  • the genetic locus comprising the genes listed in Table 2 may be referred to as the "gene”, “nucleic acid”, “locus”, “genetic locus” or “polynucleotide”. Each refers to polynucleotides, all of which are in the gene region including its 5' or 3' terminal regions, promoter, introns or exons. Accordingly, the genes of the present invention are intended to include coding sequences, intervening sequences and regulatory elements controlling transcription and/or translation. A genetic locus is intended to include all allelic variations of the DNA sequence on either or both chromosomes. Consequently, homozygous and heterozygous variations of the instant genetic loci are contemplated herein.
  • the present invention provides a genetic panel comprising different profiles of genes or mutations therein for different neurological conditions.
  • profiles include polymorphisms, although any nucleotide substitution, addition, deletion or insertion or other mutation in one or more genetic loci is encompassed by the present invention when associated with a neurological condition.
  • the present invention extends to rare mutations which although not present in larger numbers of individuals in a population, when the mutation is present in combination with at least one other mutation, it leads to a verifiable association between a responder or non-responder to a drug.
  • the present invention is not to be limited to all the genes in the genetic panel but rather two or more genes in Table 2. Particular genes and mutations of interest are listed in Tables 2A 1 and 2A ⁇ .
  • polymorphism refers to a difference in a DNA or RNA sequence or sequences among individuals, groups or populations which give rise to a statistically significant treatment outcome.
  • genetic polymorphisms include mutations that result by chance, induced by external features or are inherited.
  • nucleotide changes contemplated herein include single nucleotide polymorphisms (SNPs), multi-nucleotide polymorphisms (MNPs), frame shift mutations, including insertions and deletions (also called deletion insertion polymorphisms or DIPS), nucleotide substitutions, nonsense mutations, rearrangements and microsatellites.
  • Two or more polymorphisms may also be used either at the same allele (i.e. haplotypes) or at different alleles. All these mutations are encompassed by the term "polymorphism”.
  • Neurological conditions include, psychiatric and psychological conditions, phenotypes and states. Examples contemplated by the present invention include conditions related to dopamine pathway function and the function of associated neurotransmitters
  • GABA glutamate
  • serotonin including but are not limited to epilepsy, addiction, dementia, anxiety disorders, bipolar disorder, schizophrenia, Tourette's syndrome, obsessive compulsive disorder (OCD), panic disorder, PTSD, phobias, acute stress disorder, adjustment disorder, agoraphobia without history of panic disorder, alcohol dependence
  • amphetamine dependence brief psychotic disorder, cannabis dependence, cocaine dependence, cyclothymic disorder, delirium, delusional disorder, dysthymic disorder, generalized anxiety disorder, hallucinogen dependence, major depressive disorder, nicotine dependence, opioid dependence, paranoid personality disorder, Parkinson's disease, schizoaffective disorder, schizoid personality disorder, schizophreniform disorder, schizotypal personality disorder, sedative dependence, shared psychotic disorder, smoking dependence and social phobia.
  • One particular example of a neurological condition is epilepsy or a related disorder.
  • Two or more mutations in the genes in Table 2 can predict a treatment outcome for epilepsy or its related disorder.
  • Particular examples of mutations are set forth in Tables 2B through 21.
  • the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs391 1833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLCl A3 rs4869682
  • a further aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLCl A3
  • rs3911833 r
  • Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLC1A3 rs4869682
  • a further aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Still yet another aspect of the present invention provides a method for a genotype- based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs67826 both in SCN4B
  • rs4869682 SLC1A3
  • rs3911833
  • a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs2283170
  • rs658624 and rs678262 both in SCN4B
  • SLCl A3 rs4869682
  • Another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • Still another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
  • GBR2 also known as GPR51
  • KCNQl rs658624 and rs67826
  • SLC1A3 rs3911833
  • KNC1/MY0D1 rs507450
  • GRIA4 rsl82623
  • reference to a "positive” or “adverse” treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
  • the present invention further contemplates a method for identifying a genetic profile in a subject or group of subjects associated with the likelihood of a successful therapeutic outcome or otherwise to epilepsy or a related condition, the method comprising screening individuals for two or more polymorphisms including a mutation in a gene selected from the list in Table 2, including its 5' and 3' terminal regions, promoter, introns and exons which has a statistically significant linkage or association to a therapeutic outcome.
  • the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GST A4).
  • GBR2 also known as GPR51
  • rs2283170 KCNQl
  • the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4). Any selection of genes or mutations may
  • one or more of the above SNPs may be detected in combination with MDRl (rs 1045642).
  • the genetic test may be part of an overall diagnostic protocol involving clinical assessment and the diagnostic tools. Consequently, this aspect of the present invention may be considered as part of a therapeutic protocol.
  • Reference herein to an "individual” or a “subject” includes a human which may also be considered a patient, host, recipient or target. As indicated above, the present invention extends to veterinary applications.
  • the present invention enables, therefore, a stratification of individuals based on a genetic profile.
  • the stratification or profiling enables a prediction of which treatment is likely to be most successful or appropriate or result in less recurrence or reduced adverse drug reaction.
  • SSCP single-stranded conformation polymorphism assay
  • CDGE clamped denaturing gel electrophoresis
  • HA heteroduplex analysis
  • CMC chemical mismatch cleavage
  • an allele-specific detection approach such as allele-specific oligonucleotide (ASO) hybridization can be utilized to rapidly screen large numbers of other samples for that same mutation.
  • ASO allele-specific oligonucleotide
  • Such a technique can utilize probes which are labeled with gold nanoparticles or any other reporter molecule to yield a visual color result (Elghanian et al, Science 277:1078-1081, 1997).
  • a rapid preliminary analysis to detect polymorphisms in DNA sequences can be performed by looking at a series of Southern blots of DNA cut with one or more restriction enzymes, preferably with a large number of restriction enzymes.
  • Each blot contains a series of normal individuals and a series of individuals having neurologic or neuropsychiatric diseases or disorders or any other neurological, psychiatric or psychological condition, phenotype or state.
  • Southern blots displaying hybridizing fragments (differing in length from control DNA when probed with sequences near or to the genetic locus being tested) indicate a possible mutation or polymorphism. If restriction enzymes which produce very large restriction fragments are used, then pulsed field gel electrophoresis (PFGE) is employed.
  • PFGE pulsed field gel electrophoresis
  • the desired region of the genetic locus being tested can be amplified, the resulting amplified products can be cut with a restriction enzyme and the size of fragments produced for the different polymorphisms can be determined.
  • Detection of point mutations may be accomplished by molecular cloning of the target genes and sequencing the alleles using techniques well known in the art.
  • the gene or portions of the gene may be amplified, e.g., by PCR or other amplification technique, and the amplified gene or amplified portions of the gene may be sequenced.
  • real-time PCR such as the allele specific kinetic real-time PCR assay can be used or allele specific real-time TaqMan probes.
  • primers are used which hybridize at their 3' ends to a particular target genetic locus or mutation. If the particular polymorphism or mutation is not present, an amplification product is not observed.
  • Amplification Refractory Mutation System (ARMS) can also be used, as disclosed in European Patent Application Publication No. 0332435. Insertions and deletions of genes can also be detected by cloning, sequencing and amplification.
  • RFLP restriction fragment length polymorphism
  • Such a method is particularly useful for screening relatives of an affected individual for the presence of the mutation found in that individual.
  • Other techniques for detecting insertions and deletions as known in the art can be used.
  • an oligonucleotide is designed which detects a specific sequence, and the assay is performed by detecting the presence or absence of a hybridization signal
  • the protein binds only to sequences that contain a nucleotide mismatch in a heteroduplex between mutant and wild-type sequences.
  • Mismatches are hybridized nucleic acid duplexes in which the two strands are not 100% complementary. Lack of total homology may be due to deletions, insertions, inversions or substitutions. Mismatch detection can be used to detect point mutations in the gene or in its mRNA product.
  • mismatch cleavage technique is the RNase protection method.
  • the method involves the use of a labeled riboprobe which is complementary to the human wild-type genes (i.e. such as those listed in Table 2).
  • the riboprobe and either mRNA or DNA isolated from the person are annealed (hybridized) together and subsequently digested with the enzyme RNase A which is able to detect some mismatches in a duplex RNA structure. If a mismatch is detected by RNase A, it cleaves at the site of the mismatch.
  • RNA product when the annealed RNA preparation is separated on an electrophoretic gel matrix, if a mismatch has been detected and cleaved by RNase A, an RNA product will be seen which is smaller than the full length duplex RNA for the riboprobe and the mRNA or DNA.
  • the riboprobe need not be the full length of the mRNA or gene but can be a segment of either. If the riboprobe comprises only a segment of the mRNA or gene, it will be desirable to use a number of these probes to screen the whole mRNA sequence for mismatches.
  • DNA probes can be used to detect mismatches, through enzymatic or chemical cleavage (see, for example, Cotton et al, Proc. Natl. Acad. Sci. USA 57:4033-4037, 1988; Shenk et al, Proc. Natl. Acad. Sd. USA 72:989-993, 1975; Novack et al, Proc. Natl. Acad. Sd. USA 55:586-590, 1986).
  • mismatches can be detected by shifts in the electrophoretic mobility of mismatched duplexes relative to matched duplexes (see, for example, Cariello Am. J. Human Genetics 42:726-734, 1988).
  • the cellular mRNA or DNA which might contain a mutation can be amplified using PCR (see below) before hybridization. Changes in DNA of the associated genetic polymorphisms or genetic loci can also be detected using Southern blot hybridization, especially if the changes are gross rearrangements, such as deletions and insertions. [0089] Once the site containing the polymorphisms has been amplified, the SNPs can also be detected by primer extension. Here a primer is annealed immediately adjacent to the variant site, and the 5' end is extended a single base pair by incubation with di- deoxytrinucleotides.
  • Whether the extended base was a A, T, G or C can then be determined by mass spectrometry (MALDI-TOF) or fluorescent flow cytometric analysis (Taylor et al, Biotechniques 30:661-669, 2001) or other techniques.
  • MALDI-TOF mass spectrometry
  • fluorescent flow cytometric analysis Taylor et al, Biotechniques 30:661-669, 2001
  • Nucleic acid analysis via microchip technology is also applicable to the present invention.
  • thousands of distinct oligonucleotide probes are built up in an array on a silicon chip.
  • Nucleic acids to be analyzed are fluorescently labeled and hybridized to the probes on the chip. It is also possible to study nucleic acid-protein interactions using these nucleic acid microchips.
  • the method is one of parallel processing of many, including thousands, of probes at once and can tremendously increase the rate of analysis.
  • Mutations falling outside the coding region of the target loci can be detected by examining the non-coding regions, such as introns and regulatory sequences near or within the genes.
  • non-coding regions such as introns and regulatory sequences near or within the genes.
  • Alteration of mRNA expression from the genetic loci can be detected by any techniques known in the art. These include Northern blot analysis, PCR amplification and
  • RNA expression indicates an alteration of the wild-type gene.
  • Alteration of wild-type genes can also be detected by screening for alteration of wild-type protein.
  • monoclonal antibodies immunoreactive with a target protein i.e. two or more proteins encoded by one or more genes listed in Table 2
  • a target protein i.e. two or more proteins encoded by one or more genes listed in Table 2
  • Antibodies specific for products of mutant alleles could also be used to detect mutant gene product.
  • Such immunological assays can be done in any convenient formats known in the art. These include Western blots, immunohistochemical assays and ELISA assays.
  • Any means for detecting an altered protein can be used to detect alteration of the wild-type protein.
  • Functional assays such as protein binding determinations, can be used.
  • assays can be used which detect the protein biochemical function. Finding a mutant gene product indicates alteration of a wild-type gene product.
  • the present invention further extends to a method for identifying a genetic basis behind a successful or adverse treatment protocol for a neurological condition in an individual, the method comprising obtaining a biological sample from the individual and detecting two or more mutations in one or more proteins encoded by one or more genes listed in Table 2.
  • the neurological condition is epilepsy or a related condition.
  • the altered amino acid sequence may be detected via specific antibodies which can discriminate between the presence or absence of an amino acid change, by amino acid sequencing, by a change in protein activity or cell phenotype and/or via the presence of particular metabolites if the protein is associated with a biochemical pathway.
  • a mutant gene or corresponding gene products can also be detected in other human body samples which contain DNA, such as serum, stool, urine and sputum.
  • DNA such as serum, stool, urine and sputum.
  • the same techniques discussed above for detection of mutant genes or gene products in tissues can be applied to other body samples. By screening such body samples, an early determination can be achieved for subjects on a particular drug or about to be prescribed a particular drug.
  • the present invention extends to two or more isolated oligonucleotides which comprise from about three to about 1000 consecutive nucleotides from the gene or its corresponding cDNA or mRNA as listed in Table 2 which encompass at least two polymorphisms or mutations associated with a particular therapeutic outcome for a neurological condition.
  • the neurological condition is epilepsy or a related condition.
  • one of the at least two primers is involved in an amplification reaction to amplify a target sequence. If this primer is also labeled with a reporter molecule, the amplification reaction will result in the incorporation of any of the label into the amplified product.
  • amplification product and “amplicon” may be used interchangeably.
  • primers and the amplicons of the present invention may also be modified in a manner which provides either a detectable signal or aids in the purification of the amplified product.
  • a range of labels providing a detectable signal may be employed.
  • the label may be associated with a primer or amplicon or it may be attached to an intermediate which subsequently binds to the primer or amplicon.
  • the label may be selected from a group including a chromogen, a catalyst, an enzyme, a fluorophore, a luminescent molecule, a chemiluminescent molecule, a lanthanide ion such as Europium (Eu 34 ), a radioisotope and a direct visual label.
  • a colloidal metallic or non-metallic particular a dye particle, an enzyme or a substrate, an organic polymer, a latex particle, a liposome, or other vesicle containing a signal producing substance and the like.
  • Suitable enzyme labels useful in the present invention include alkaline phosphatase, horseradish peroxidase, luciferase, ⁇ - galactosidase, glucose oxidase, lysozyme, malate dehydrogenase and the like.
  • the enzyme label may be used alone or in combination with a second enzyme which is in solution.
  • a fluorophore which may be used as a suitable label in accordance with the present invention includes, but is not limited to, fluorescein-isothiocyanate (FITC), and the fluorochrome is selected from FITC, cyanine-2, Cyanine-3, Cyanine-3.5, Cyanine-5, Cyanine-7, fluorescein, Texas red, rhodamine, lissamine and phycoerythrin.
  • the primers or amplicons may additionally be incorporated on a bead or other solid support.
  • a biological sample such as blood is obtained and analyzed for the presence or absence of a panel of target alleles comprising from about two to 100 alleles or from about two to 50 alleles or from two to about 30 alleles of the genetic loci identified as being statistically significantly associated with the treatment outcome for epilepsy. Results of these tests and interpretive information are returned to the health care provider a decision on which medicament is appropriate.
  • diagnoses may be performed by diagnostic laboratories, or, alternatively, diagnostic kits are manufactured and sold to health care providers or to private individuals for self-diagnosis. Suitable diagnostic techniques include those described herein as well as those described in US Patent Numbers 5,837,492; 5,800,998 and 5,891,628.
  • the candidate gene approach is ideal as unrelated cases and controls suffice. Although the whole genome approach is perhaps a better model to base the study on, it has financial limitation, and would cost a substantial amount more to organize and employ. Additionally, the candidate gene approach is appropriate as there is a significant understanding of the pathophysiological of epilepsy. A common concern with this approach is the lack in knowledge of functional variants, and this is why the study incorporates tagger SNPs that have no current functional significance.
  • SNPs may not the casual variants, but may lie within or near the casual variant by virtue of linkage disequilibrium.
  • Table 2 lists the candidate genes used in this study.
  • SNP selection strategy [0108] All common variation across the list of candidate genes in Table 2 were examined through a combined direct (genotyped) and indirect (captured via linkage disequilibrium) mapping strategy. Functional variation was assessed directly whilst all other variation of unknown function was assessed indirectly through tagger SNPs. Additionally, 100 SNPs were incorporated for stratification analysis; these SNPs are neutral with respect to epilepsy (based on current available literature) and spaced far apart so as to be in linkage equilibrium. A detailed description of the protocol used to select candidate genes, and furthermore, SNPs to be used is provided below.
  • SNP selection was by a fuse a tagging (map based) with a functional (sequence based) approach to detect variation functional in the development and treatment of common epilepsy treatment outcomes; using a candidate gene approach with selection of genes based on biology.
  • DNA was extracted and purified from the blood samples of the first 179 patients who had at least one year follow-up and a technically adequate blood sample available. This DNA was genotyped for 4,041 SNPs from a pre-selected set of 279 candidate genes. Noise reduction was performed by removing SNPs with missing values or little variation amongst treatment outcomes; improving the quality of the available dataset.
  • Non-responsiveness was defined as spontaneous seizures to initial drug therapy where the patient had reached therapeutic levels. In instances where a patient was taken off an initial AED before pharmacoresponsiveness could be determined by ADRs, pharmacoresponsiveness was determined on a subsequent AED used.
  • ADRs were phenotyped by a clinical neurologist. Reported ADRs were then clustered into five clinically relevant groups: neurological, metabolic, immunological, gastrointestinal (GIT), and other. This clustering provides an opportunity for deeper analysis by increasing the number of events within each general ADR group.
  • Table 3 displays all patients from the genotyped cohort who experienced an ADR, which ADR, and the general class of ADRs it falls into.
  • GOW Gain of Weight; >5kg within first 3 months OR >7.5kg over 2 year course
  • GIT has two sub-groups: Nausea and Diarrhoea Other includes: Hair loss and Acne Statistical Methods
  • the samples were analyzed by four different approaches: two univeriate (Fishers t- tests and chi-square, and using SAS Genetic Marker) and two multivariate (GeneRave and Hybrid approach). The analyses were completed on two treatment outcomes, responsiveness and ADRs. Once as a combined cohort then further subdivided based on the two most commonly prescribed AEDs carbamazepine and valproate.
  • the PROC %TPLOT function is a useful application in detecting blocks of LD and is methodologically similar to Haploview. It combines output from the CASECONTROL, PSMOOTH, and HAPLOTYPE procedures. It allows the visualization of smoothed p- values from Hardy Weinberg Equilibrium (HWE) tests, tests for linkage disequilibrium between SNPs and association tests between SNPs and pharmacoresi stance.
  • HWE Hardy Weinberg Equilibrium
  • GeneRave's RChip is an example of the many commercially available suites.
  • the RChip suite was selected for its ability to analyse the relationship between response variables and a set of predictors when the number of predictors far exceeds the number of observations. This suite was designed based on the leukaemia data published by Yeoh et al, Cancer CU 7:133-143, 2002.
  • RChip has the ability to identify genes/SNPs that discriminate between different phenotypes and this capability was tested using the epilepsy treatment outcome data.
  • a predictor can be developed that classifies 'test data', and in doing so determines the features (SNPs) responsible for the classification.
  • the aim being to find SNPs that based on presence can discriminate between two categorical classes.
  • a scoring system is required to reflect the SNPs influence on the distinction between classes.
  • the score adopted in this analysis is the signal-to-noise statistic (golub score), described more fully below.
  • the classifier is built in 2 steps: First, the most differentially expressed SNPs between the classes are selected using the golub score. Second, the £NN classifier evaluates the number of signature SNPs (N) and the number of nearest neighbours (k) to use by optimising the classification performance on validation datasets.
  • the pseudo-code for the classifier program which I built using PERL programming language is provided herein under:
  • Computer then creates a hash reference that reads in the training data and then another hash reference for the validation data.
  • Computer then stores the integer value for the number of entries in the training set. • The computer then takes each validation entry and checks it by the corresponding training entry to calculate the Euclidean distance between the training and the validation, and then pushes this into the array @calc.
  • @calc now holds the comparison for each training set with the validation set, and sorts these in ascending order.
  • the computer loops through @calc and counts the number of treatment outcome 1 occurrences, and the number of treatment outcome 2 occurrences from the training data for that one validation kernel.
  • the £NN classifier requires that there be non-null numbers for any of the three genotypes in any SNP. This includes when sub-dividing cohort into 5 sub-cohorts. Therefore SNPs that had multiple missing values or patients with multiple missing SNPs were removed from the analysis.
  • the crux of the &NN classifier is that the class of a validation sample is decided by the majority class among its k nearest neighbours. A neighbour is deemed nearest if it has the smallest Euclidean distance in the ⁇ -dimensional space, where N is the number of top ranked SNPs chosen.
  • the algorithm calculates the distance from validation sample y to each training sample x, using Euclidean distance. Then the researcher identifies the k training samples with the smallest distance to y, and checks what the majority class is amongst them. This is re-done for each validation sample. To avoid a tied vote, k is chosen as an odd number.
  • the &NN algorithm uses all attributes in the training set and plots them into the data space.
  • a kernel is formed in data space centred on the validation case. This kernel is hypersphere shaped and is just large enough to contain the k nearest neighbours of the validation case using a Euclidean distance metric. Classification is then performed according to the k nearest neighbours found in the kernel.
  • the most important success criterion of this approach is the evaluation of the classifiers performance.
  • Two strategies were trialled, the sturdy Training ⁇ Validation ⁇ Test method, which is referred to as the TVT method, and the cross-validation method. Initially the TVT method was applied; however this method consumed many cases in the validation and test sets, leaving the training set, which is the classification building step with a low number of cases to build the classifier upon. Due to this size limitation, the alternative cross-validation approach was adopted. The advantage is that all the data are used for cross-training and testing, whilst the validation remains completely independent to the training.
  • V was chosen to be five cohorts in this analysis. Reasoning for five cohorts: getting the optimal training (classification building) dataset as more individuals would be used in building the classifier. Thus, by dividing the cohort of 119 patients into five groups it allows for a training dataset of approximately 100 patients on which to build the classifier and on average 23 in the validation cohorts. Ideally larger validation sets would cater for any noise that may affect the classification, however due to dataset limitations, 23 individuals in the validation set was the target number taking all aspects into consideration.
  • the top 15 ranked SNPs for each of the V test sample estimates were also recorded. These SNPs were used as a cross-check to determine which five SNPs reappear in the top 15 across the V test sample estimates. Once the top five SNPs were selected, classification was re-run for each validation cohort using the combined top five SNPs. Fishers t-tests were then performed for each validation cohort to determine the strength of the classifier for each.
  • the 49 patients are independent of the initial 179 patients that were genotyped. They were passed through the classifier and also SAS Genetics Marker code. Regardless of model derivation, a significant result in this replication would suffice to prove the effect that the 5 SNPs used in the classifier have in predicting pharmacoresponsiveness.
  • the classifier and SNPs were tested on a population of community recruited 189 chronic treated epilepsy patients. Initial phenotype data included patients without a follow up (21), patients without a recorded AED (22), patients without recorded seizure (15), patients with date errors (23), were removed leaving final analysis of 108 patients. These 108 patients were used to validate classifiers performance in a separate population from which the classifier was built, and the 8 SNPs were looked at individually.
  • the original model was developed with the flexibility of including two-year follow-ups when deciding responsiveness, the chronic cohort required a much stricter 12 month follow-up restriction.
  • Previous studies of pharmacoresistance in chronic epilepsy populations have set the responder criterion to less than 3 seizures within the year. Flexibility is allowed for in the restrictions to patients who have been seizure free within 12 months of their interview. 12 months of seizure freedom is required for obtaining a motor-vehicle license in Australia.
  • 121 patients with 12 months follow-up were analyzed.
  • the top ranked SNP was from the GABBR2 also known as the GPR51 gene (rs2808526) with a p-value of 0.0006 and a genotypic OR of 21 (Table 7).
  • a Average p-value is the average value of the 100 control runs b
  • the real result indicates the actual top p-value obtained for the GABBR2 rs2808526 SNP c
  • the top SNP was from the GLUL gene with a p-value of 0.07 and an OR of 9.2. This decrease in significance implies it has an effect but is unlikely to be drug specific.
  • rs2808526 from the GABBR2 gene (also known as GPR51) was the top SNP with a genotypic probability of 0.0001 and a genotypic chi-squared value of 18.25. Furthermore, the allelic probability was highly significant at 4.19E-05 and chi-square of 16.8, and the trend probability was 2.88E-05 with a chi-squared value of 17.5.
  • Other SNPs that appeared highly significant across the three categories of tests were rs2229944 from the GABRB2 gene, rs507450 from the GRIA4 gene, and rs658624 from the SCN4B gene. The collection of top five SNPs is displayed in Table 9.
  • TPLOT has its capacity to bunch SNPs in linkage disequilibrium together. This is advantageous because the number of linkage blocks can be used as a better estimate in multiple hypothesis correction than individual SNPs.
  • the output required to observe collinearity between SNPs in the large epilepsy dataset in this project is 8,166,861. Therefore, neither TPLOT nor Haploview has the capacity to analyze such large datasets.
  • rs658624 and rs2808526 were found to be present in all five cohorts; rs4869682 was present in cohorts 1-4; rs3911833 was present in cohorts 1,3,4,5; rs678262 was present in cohorts 1,3,5; rs6001641 was present in cohorts 1,3,4; rs7153926 was present in cohorts 2,3,5; and rs2283170 was present in cohorts 3,4,5.
  • rs6001641, rs3911833 and rs7153926 were removed for their low frequencies across genotypes, leaving five SNPs (rs658624 [SCN4B], rs2808526 [GABBR2 (GPR51)], rs2283170 [KCNQl], rs4869482 [SLCl A3] and rs678262 [SCN4B ⁇ ).
  • the k nearest number was chosen to be nine as nine neighbours produced the best detection rate in the cross-validated samples.
  • Each validation cohort was then re-run through the classifier with these five SNPs exclusively.
  • each cohort produced an improved level of prediction compared to the original best five SNPs from 4,041. This indicates selection of these five SNPs has removed noise from other SNPs which were initially ranked as the top five for each cohort.
  • Table 11 reports the Fisher t- test scores plus sensitivity, specificity and predictive value tests to measure classifier accuracy for each of the validation cohort predictors.
  • TP Number of true positives: Responders correctly classified as responders.
  • b FP (type I error) Number of false positives: Non-responders incorrectly classified as responders.
  • c TN Number of true negatives: Non-responders correctly classified as non- responders.
  • FN (Type II error) Number of false negatives: Responders incorrectly classified as non-responders
  • a classification tree is constructed using training set (L- L v ). e.g. L] is classified based on training set (L 2 +L 3 +L 4 +L 5 ).
  • Drug specific pharmacoresponsiveness [0160] The 115 patients used in constructing the pharmacoresponsiveness classifier were sub-categorised into two drug types, carbamazepine and valproate. The number of patients remaining was insufficient to build a reliable classifier, with no cross-validating SNPs observed.
  • the optimized training set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3).
  • Another set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
  • the optimized training set provides 74% sensitivity and 52% specificity, with a p- value of 0.005 for predicting responsiveness.
  • Adverse drug reactions occur in approximately 40% of patients commencing treatment with an anti-epileptic drug and can affect the neurological, immunological, metabolic or gastrointestinal system.
  • the most common ADRs are: sedation, neurocognitive effects (especially poor concentration and memory), ataxia, weight gain, skin rash, bone mineral density loss and increase fracture risk.
  • the univariable/multivariate classification system is used aimed at validating a correlation between ADR and SNPs or other mutations.
  • the mutations may be in one or more genes listed in Table 2.
  • clinical samples are collected from a sufficient number of patients so that each ADR subgroup is represented by a n times wherein n is at least 1 ensuring that a statistically significant threshold is reached after application of the model.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Health & Medical Sciences (AREA)
  • Organic Chemistry (AREA)
  • Wood Science & Technology (AREA)
  • Analytical Chemistry (AREA)
  • Zoology (AREA)
  • Genetics & Genomics (AREA)
  • Engineering & Computer Science (AREA)
  • Pathology (AREA)
  • Immunology (AREA)
  • Microbiology (AREA)
  • Molecular Biology (AREA)
  • Biotechnology (AREA)
  • Biophysics (AREA)
  • Physics & Mathematics (AREA)
  • Biochemistry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

The present invention relates generally to diagnostic assays, therapeutic protocols and medicinal predictor model validation. More particularly, genetic tests are provided which determine the suitability of a medicament in the treatment or prophylaxis of a neurological condition. Even more particularly, the present invention provides genetic assays to measure the potential for, or likelihood of, a subject having a positive or adverse treatment response to a neurological medicament. The present invention is particularly useful in the practice of personalized medicine.

Description

A DIAGNOSTIC ASSAY
FILING DATA
[0001] This application is associated with and claims priority from Australian Provisional Application No. 2007905783, filed on 22 October, 2007 and Australian Provisional Application No. 2007906200, filed on 12 November, 2007, the entire contents of which are incorporated herein by reference.
FIELD
[0002] The present invention relates generally to diagnostic assays, therapeutic protocols and medicinal predictor model validation. More particularly, genetic tests are provided which determine the suitability of a medicament in the treatment or prophylaxis of a neurological condition. Even more particularly, the present invention provides genetic assays to measure the potential for, or likelihood of, a subject having a positive or adverse treatment response to a neurological medicament. The present invention is particularly useful in the practice of personalized medicine.
BACKGROUND
[0003] Bibliographic details of the publications referred to by author in this specification are collected at the end of the description.
[0004] Reference to any prior art in this specification is not, and should not be taken as, an acknowledgment or any form of suggestion that this prior art forms part of the common general knowledge in any country.
[0005] Epilepsy is estimated to affect approximately 50 million people globally according to statistics provided by the World Health Organization (WHO; 1995). Current epilepsy treatment is less than satisfactory, with 40% of patients having a significant adverse drug reaction (ADR) and 20-40% experiencing seizure recurrence (Mattson et al, The New England Journal of Medicine 575:145-151, 1985; Kwan and Brodie, The New England Journal of Medicine 342:314-319, 2000). Importantly, there is significant inter-individual variability in response to an anti-epileptic drug (AED), and also intra-individual variability in response between different drugs. Currently, the choice of AEDs prescribed is essentially on a "trial and error" basis, with little prognostic useful information being available to the clinician to guide which drugs the patient is likely to positively respond.
[0006] Locating genetic predictors of individual drug response (both seizure control and ADRs) to AED treatment would significantly improve the current practice of AED prescription. The ability to predict an individual patient's response to a medication before prescribing the drug would not only provide significant personal benefits for individuals, but also a substantial economic benefit to both government health schemes and the community, by reducing the cost of ineffective and dangerous treatments. The significance of this is illustrated by an analysis of studies in which the cause of hospitalization was determined. This study found that approximately 1.5 million hospitalizations a year were caused by ADRs (not exclusive to epilepsy patients) [Lazarou et al, JAMA 279:1200-1205, 1998]. Additionally, studies estimate that uncontrolled seizures cost US$2,250 to $3,205 per year per patient (Begley and Beghi, Epilepsia 41:342-219, 2002). Uncontrolled seizures also have includes indirect costs such as loss of productivity and opportunity which account for 70-85% of total disease-related costs (Akobundu et al, PharmacoEconomics 24:869-890, 2006).
[0007] The occurrence of ADRs and pharmacoresistance to AED medication is affected by the action of multiple genes, including those involved in the metabolic pathway, mode of adsorption, transportation and receptors of the AEDs, as well as immunological processes. The principal behind pharmacogenetics is that mutations within these genes could lead to their malfunction by interfering with their effect on, or of, the AED resulting in an altered function (pharmacoresistance or level of pharmacosensitivity) or lead to increased blood levels of these drugs or enhanced unwanted pharmacodynamic or immunological responses, predisposing the individuals to a greater risk of developing ADRs.
[0008] In work leading up to the present invention, candidate genes which mediate the aforementioned outcomes are identified from the pharmacokinetics and pharmacodynamics of AEDs and genes involved in the pathophysiology of epilepsies.
[0009] These genes are used to determine the relationship between genotype and neurological disease outcome or potential outcome in response to medication.
SUMMARY
[0010] Throughout this specification, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element or integer or group of elements or integers but not the exclusion of any other element or integer or group of elements or integers.
[0011] The present invention employs a pharmacogenomic approach to screen datasets of two or more genetic markers, such as genes, and/or two or more single nucleotide polymorphisms (SNPs) in genes as a predictor of the likelihood or otherwise that a subject will favorably respond to a drug used to treat a neurological condition. Pharmacoresistance or pharmacosensitivity to a particular medicament used for a neurological disease is proposed herein to be predictable based on the presence or absence of a dataset of SNPs in selected genes. Even more particularly, the neurological disease is epilepsy or a related condition. Hence, the present invention provides a pharmacogenomics approach in the assessment of therapeutic outcome potential or pharmacoresistance or pharmacosensitivity to a particular neurological drug. Reference to "pharmacogenomics" in this context includes the study of a spectrum of genes which potentially influences a drug response in a subject.
[0012] The present invention employs a multivariate approach to pharmacoresistance determination and assessment. The multivariate approach is a predictor of a combinatory genetic effect of drug response. Hence, the present invention provides a method for generating a validated medicinal predictor model for use in personalized medicine and epidemiological studies of population groups.
[0013] The diagnostic assay of the present invention enables a practitioner to select a particular drug or avoid a drug to ensure a reduced likelihood of development of an adverse drug reaction (ADR) in a subject. This leads to greater opportunity for successful treatment, improved quality of life to the subject and reduced health costs. It also enables the practitioner to predict, with an increased degree of certainty, the chance that a patient will have recurrent seizures or other neurological symptoms despite the commencement of drug treatment. This has important implications in providing advice to a patient in relation to daily activities.
[0014] Accordingly, one aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
[0015] A further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament. In this context, reference to a "positive" or "adverse" treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
[0016] Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
[0017] Reference to a "neurological condition" includes a neurological, psychiatric or psychological condition, phenotype or state including a sub-threshold neurological, psychiatric or psychological condition, phenotype or state. In one embodiment, the neurological condition is epilepsy or a related condition.
[0018] Generally, the subject is a human. However, the present invention has veterinary applications in relation to non-human animals. Reference to a "mutation" generally includes a nucleotide polymorphism such as a single nucleotide polymorphism (SNP) and a multi-nucleotide polymorphism (MNP).
[0019] Datasets of genes or mutations in genes are set forth in Tables 2 A through 21. Reference to "Table 2" includes any or all of Tables 2A through 21. Table 2A includes Table 2A1 and Table 2A11. These lists of genes and mutations represent a first knowledge base. The correlation or weighting between selected mutations in selected genes to a pharmacogenomic effect on medication represents a second knowledge base. The correlation may be determined via an algorithm which represents a training tool.
[0020] Hence, the present invention contemplates use of a set of genes in a first knowledge base to correlate mutations in the form of a second knowledge base via a training tool (i.e. hybrid classifier system) with pharmacosensitivity, pharmacoresistance or ADR to a drug used in neurological treatment. A subset of genes or mutations in genes is selected based on the drug employed. It is proposed herein that combinations of nucleotide mutations such as SNPs is collectively more highly predictive of a responder compared to univariate analysis based on a single mutation.
[0021] Accordingly, one aspect of the present invention provides a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, said method comprising screening for the presence of two or more mutations in genes selected from GABBR2, KCNQl, SCN4B and SLC1A3 which mutations correlate to potential responsiveness of the subject to the medicament.
[0022] In a further embodiment, mutations in one or more of KCNCl /MYODl, GRIA4 and/or GSTA4 are screened in addition to one or more of GABBR2, KCNQl, SCN4B and/or SLCl A3.
[0023] Most particularly, the SNPs contemplated herein comprise two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3).
[0024] In another embodiment, the SNPs contemplated herein are one or more of rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) in combination with one or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3)..
[0025] In a further embodiment, the SNPs contemplated herein are two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4).
[0026] In yet a further embodiment, one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
[0027] Reference to "two or more" in relation to the latter aspects of the invention includes 2, 3, 4, 5, 6, 7 or 8 SNPs. Reference to one or more includes 1, 2, 3, 4, 5, 6, 7 or 8 SNPs.
[0028] The second knowledge base may be further particularized into an optimized training set. Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs in genes selected from the list comprising rs2808526
(GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170
(KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833
(KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4) which SNPs correlate to potential responsiveness of the subject to the medicament.
[0029] A further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament. As indicated above, reference to a "positive" or "adverse" treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
[0030] Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0031] Kits, computer programs and treatment and diagnostic protocols are also encompassed by the present invention. A treatment protocol comprises detecting or predicting that a subject will likely be a responder to a particular drug or panel of drugs and then selecting the appropriate drug. A "computer program" includes web-based assays where SNP information is provided to a web site which determines the likelihood of pharmacosensitivity, pharmacoresistance or an ADR.
[0032] A business method is therefore provided comprising inputting into a web-based site information concerning the presence of two or more mutations in an optimized training set of genes selected from Table 2 wherein the web-based site provides an interactive response providing information on potential pharmacosensitivity or pharmacoresistance or an ADR to a drug used in neurological treatment. Examples of mutations in genes includes two or more ofrs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
[0033] In yet a further embodiment, one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
[0034] Terms such as "drug" and "medicament" may be used interchangeably throughout the subject specification and includes in one embodiment an anti-epileptic drug (AED) or class of AED.
[0035] Abbreviations used herein are defined in Table 1.
Table 1 Abbreviations
Figure imgf000011_0001
Table : 2
Candidate genes
Gene name(a) ref seq(b) gene status(c) Chr(d) Strand(e) Genomic address'0
GPR51
(GABBR2) NM 005458 intronless 9 - chr9:98128920-98561034
KCNQl NM 000218 intronless 11 + chrl 1:2412796-2827915
SCN4B NM 174934 full 11 - chrl 1:117508303-117538745
SLC1A3 NM_004172 full 5 + chr5:36632447-36725191
KCNCl
(MYODl) NM 004976 full 11 + chrl l:17704121-17751752
GRIAl NM 000827 intronless 5 + chr5:152840498-153172354
GSTA4 NM 001512 full 6 - chr6:52949709-52978099
ABAT NM 020686 intronless 16 + chrl6:8665945-8786932
ABCA2 NM 001606 func only 9 - chr9: 137176522-137208577
ABCBl NM 000927 func only 7 - chr7:86776598-86997215
ABCCl NM 004996 func only 16 + chrl6:15940934-16144774
ABCClO NM 033450 func only 6 + chr6:43497466-43527141
ABCC2 NM 000392 func only 10 + chrl0: 101522562-101602571
ABCC3 NM 003786 func only 17 + chrl7:46057226-46125061
ABCC4 NM 005845 func only 13 - chrl 3.94469090-94761684
ABCC5 NM 005688 func only 3 - chr3: 185119427-185228429
ABCC6 NM 001171 func only 16 - chrl6:16150491-16234815
ABCG2 NM 004827 func only 4 - chr4:89367595-89447190
ACHE NM 015831 full 7 - chr7:100131266-100148192
ADK NM 006721 func only 10 + chrl0:75570970-76140065
ADORAl NM 000674 full 1 + chrl . -199781438-199869189
ADORA2A NM 000675 full 22 + chr22:23143645-23163878
ADORA2B NM 000676 full 17 + chrl7:15778955-15820935
ADORA3 NM 020683 full 1 - chrl :111737013-111768152
ADRB2 NM 000024 func only 5 + chr5:148176368-148189379
ALDH5A1 NM 170740 full 6 + chr6:24593175-24646413
ALPL NM 000478 full 1 + chrl :21571174-21651208
ARHGEF9 NM 015185 full X - chrX:62636868-62768014
BAKl NM 001188 full 6 - chr6:33647306-33665959
BCL2 NM 000633 intronless 18 - chrl8:58940558-59147593
BCL2L1 NM 138578 full 20 - chr20:29714923-29784317
BRD2 NM 005104 full 6 + chr6:33034414-33058059
BSND NM 057176 full 1 + chrl :55166637-55187485
CACNAlA NM 023035 intronless 19 - chrl9:13178114-13488317
CACNAlB NM 000718 full 9 + chr9: 138038077-138293468
CACNAlC NM 000719 func only 12 + chrl2:2022724-2673368
CACNAlD NM 000720 intronless 3 + chr3:53494115-53822000
CACNAlE NM 000721 intronless 1 + chrl :178174372-178503370
CACNAlF NM 005183 full X - chrX:48816894-48856204
CACNAlG NM 018896 full 17 + chrl 7:45983447-46060430
CACNAlH NM 021098 full 16 + chrl6: 1133241-1212772
CACNAlI NM 021096 full 22 + chr22:38281257-38411238
CACNAlS NM 000069 full 1 - chrl: 197739298-197823351
CACNA2D1 NM 000722 intronless 7 - chr7:81223068-81727682
CACNA2D2 NM 001005505 full 3 chr3:50374235-50525896
CACNA2D3 NM 018398 func only 3 + chr3:54121732-55084622
CACNA2D4 NM 172364 full 12 - chrl2: 1770384-1908131
CACNBl NM 000723 full 17 _ chrl7:34582234-34617427 Gene name(a) ref seq(b) gene status(c) Chr(d) Strand(e Genomic address'0
CACNB2 NM 201596 fiinc only 10 + chrlθ:18459611-18871044
CACNB3 NM 000725 full 12 + chrl2:47488778-47509991
CACNB4 NM 000726 intronless 2 - chr2: 152518649-152791052
CACNGl NM 000727 full 17 + chrl7:62461167-62484371
CACNG2 NM 006078 intronless 22 - chr22:35283603-35433403
CACNG3 NM 006539 intronless 16 + chrl 6:24164376-24282237
CACNG4 NM 014405 full 17 + chrl7:62381474-62460980
CACNG5 NM 145811 full 17 + chrl7:62293912-62312818
CACNG6 NM 145814 full 19 + chrl9:59177353-59208730
CACNG7 NM 031896 full 19 + chrl9:59097882-59139080
CACNG8 NM 031895 full 19 + chrl9:59148105-59178951
CAMK2A NM 015981 func only 5 - chr5: 149578247-149659529
CAMK2B NM 001220 func only 7 chr7:44031136-44148464
CAMK2D NM 001221 func only 4 chr4: l 14732168-115050332
CDK5 NM 004935 full 7 - chr7: 150187546- 150202644
CHAT NM 020549 full 10 + chrl 0:50482061-50544156
CHRMl NM 000738 full 11 chrl 1 :62431727-62455588
CHRM2 NM 001006630 intronless 7 + chr7: 136000653- 136160026
CHRM3 NM 000740 intronless 1 + chrl :236108413-236399756
CHRM4 NM 000741 full 11 chrl 1:46362215-46374683
CHRM5 NM 012125 full 15 + chrl5:32087719-32145579
CHRNAl NM 000079 full 2 - chr2: 175436830-175464688
CHRNAlO NM 020402 full 11 chrl 1:3642392-3659190
CHRNA2 NM 000742 full 8 chr8:27373182-27402675
CHRNA3 NM 000743 full 15 - chrl5:76673706-76710377
CHRNA4 NM 000744 full 20 - chr20:61444108-61473192
CHRNA5 NM 000745 full 15 + chrl5:76634960-76674513
CHRNA6 NM 004198 full 8 chr8:42725937-42752776
CHRNA7 NM 000746 intronless 15 + chrl5:30100017-30249525
CHRNA9 NM 017581 full 4 + chr4:40168396-40198901
CHRNBl NM 000747 full 17 + chrl7:7279129-7302655
CHRNB2 NM 000748 full 1 + chrl : 151343329-151363156
CHRNB3 NM 000749 full 8 + chr8:42661718-42712366
CHRNB4 NM 000750 full 15 chrl 5:76702690-76730642
CLCNl NM 000083 full 7 + chr7: 142520055-142566934
CLCN2 NM 004366 full 3 - chr3: 185546096-185571969
CLCN3 NM 173872 full 4 + chr4:170906451-171017886
CLCN4 NM 001830 full X + chrX:9924772-10014256
CLCN5 NM 000084 full X + chrX:49527191-49561556
CLCN6 NM 001286 full 1 + chrl : 11790559-11838451
CLCN7 NM 001287 full 16 - chrlό: 1434345-1475013
CLCNKA NM 004070 full 1 + chrl:16084415-16106850
CLCNKB NM 000085 full 1 + chrl : 16105657-16129782
CNT2 NM 004212 func only 15 + chrl5:43321725-43356425
CPLXl NM 006651 func only 4 - chr4:767575-819775
CPLX2 NM 006650 func only 5 + chr5:175146215-175244629
CSNKlAl NM 001892 func only 5 - chr5:148854037-148921200
CYP 1A2 NM 000761 full 15 + chrl5:72818236-72836994
CYP2A6 NM 000762 full 19 - chrl9:46040283-46058180
CYP2B6 NM 000767 full 19 + chrl9:46179043-46217141
CYP2C19 NM 000769 full 10 + chrl0:96502452-96603660
CYP2C9 NM 000771 full 10 + chrl0:96678429-96740137
CYP2D6 NM 000106 full 22 - chr22 :40846000-40861381
CYP2E1 NM 000773 full 10 + chrl 0:135219747- 135242501
CYP3A4 NM 017460 full 7 _ chr7:98998254-99036459 Gene name00 ref seq(b) gene status(c) Chr(d) Strand"0 Genomic address'0
CYP3A5 NM 000777 full 7 - chr7:98889468-98932257
DRD2 NM 000795 func only 11 - chrl 1 : 112784527-112861091
EFHCl NM 018100 full 6 + chr6:52383070-52466177
EPHXl NM 000120 full 1 + chr 1 :222309710-222340995
EPIM NM 194356 func only 12 - chrl2:129798027-129858691
GABARAP NM 007278 full 17 - chrl 7:7083462-7096477
GABBRl NM 001470 full 6 - chr6:29676984-29718839
GABRAl NM 000806 full 5 + chr5: 161196982-161259990
GABRA2 NM 000807 intronless 4 - chr4:46091633-46242873
GABRA3 NM 000808 full X chrX: 151006099-151300398
GABRA4 NM 000809 full 4 - chr4:46760846-46846508
GABRA5 NM 000810 intronless 15 + chrl5:24653150-24777749
GABRA6 NM 000811 full 5 + chr5: 161035547-161062689
GABRBl NM 000812 intronless 4 + chr4:46864506-47270373
GABRB2 NM 021911 intronless 5 - chr5 : 160652448- 160917708
GABRB3 NM 000814 intronless 15 - chrl 5:24338788-24579344
GABRE NM 004961 full X - chrX: 150791164-150823719
GABRGl NM 173536 full 4 - chr4:45877716-45977010
GABRG2 NM 198904 full 5 + chr5: 161417294-161516104
GABRG3 NM 033223 intronless 15 + chrl5:24789412-25452722
GABRP NM 014211 full 5 + chr5:170133342-170174626
GABRQ NM 018558 full X + chrX:151467204-151493393
GABRRl NM 002042 full 6 - chr6:89943690-89993779
GABRR2 NM 002043 full 6 - chr6:90022957-90091673
GADl NM 000817 full 2 + chr2: 171488706-171544164
GAD2 NM 000818 full 10 + chrl0:26535599-26634492
GLS NM 014905 full 2 + chr2:191561107-191656771
GLS2 NM 013267 full 12 - chrl2:55150003-55178448
GLUDl NM 005271 full 10 - chrl0:88799222-88854603
GLUD2 NM 012084 full X + chrX:l 19896996-119910329
GLUL NM 002065 full 1 - chrl :179082325-179102607
GPHN NM 020806 intronless 14 + chrl4:66033877-66719276
GRIA2 NM 000826 intronless 4 + chr4: 158489451-158645829
GRIA3 NM 007325 full X + chrX: 122033692-122349328
GRIA4 NM 000829 intronless 11 + chrl 1: 104976934-105356674
GRIKl NM 000830 intronless 21 chr21:29846736-30244153
GRIK2 NM 021956 intronless 6 + chrό: 101943674- 102624474
GRIK3 NM 000831 intronless 1 - chrl :36934706-37178937
GRIK4 NM 014619 intronless 11 + chrl 1 :120026237-120363179
GRIK5 NM 002088 full 19 - chrl9:47193312-47271797
GRINl NM 007327 full 9 + chr9: 137298678- 137340044
GRIN2A NM 000833 func only 16 chrl6:9761922-10194112
GRIN2B NM 000834 func only 12 - chrl2: 13604410-14034319
GRIN2C NM 000835 full 17 - chrl 7:70348762-70377602
GRIN2D NM 000836 full 19 + chrl9:53579943-53640205
GRIN3A NM 133445 full 9 - chr9: 101410189-101590417
GRIN3B NM 138690 full 19 + chrl9:941436-961723
GRMl NM 000838 intronless 6 + chr6: 146382111-146801427
GRM2 NM 000839 full 3 + chr3:51708003-51728663
GRM3 NM 000840 intronless 7 + chr7:85907880-86139842
GRM4 NM 000841 full 6 chr6:34096606-34219421
GRM5 NM 000842 func only 11 - chrl 1 :87879625-88430838
GRM6 NM 000843 full 5 - chr5:178336937-178364730
GRM7 NM 181874 func only 3 + chr3:6867926-7759217
GRM8 NM 000845 func only 7 - chr7: 125671607-126487261 Gene name<a) ref seq(b) gene status(c) Chr(d) Strand(e) Genomic address*0
GSTAl NM 145740 func only 6 - chr6:52765347-52766616
GSTA2 NM 000846 full 6 chr6:52722139-52746283
HCNl NM 021072 intronless 5 chr5:45296730-45741977
HCN2 NM 001194 full 19 + chrl9:530892-569157
HCN3 NM 020897 full 1 + chrl:152050446-152073711
HCN4 NM 005477 full 15 chrl5:71399987-71458230
HTR2A NM 000621 func only 13 chrl3:46304513-46378176
HTR5A NM 024012 func only 7 + chr7:154290193-154316107
KCNAl NM 000217 full 12 + chrl2:4880805-4893291
KCNAlO NM 005549 full 1 - chrl:l 10770880-110783839
KCNA2 NM 004974 full 1 chrl: l 10856818-110870387
KCNA3 NM 002232 full 1 chrl: l 10925351-110939697
KCNA4 NM 002233 full 11 chrl 1:29987341-30005064
KCNA5 NM 002234 full 12 + chrl2:5013345-5027209
KCNA6 NM 002235 full 12 + chrl2:4778602-4793839
KCNA7 NM 031886 full 19 - chrl9:54261487-54278010
KCNABl NM 172160 intronless 3 + chr3: 157311049-157740095
KCN AB2 NM 172130 full 1 + chrl :6010645-6095789
KCN AB3 NM 004732 full 17 - chrl7:7765751-7783478
KCND2 NM 012281 intronless 7 + chr7: l 19497672-119985338
KCNJlO NM 002241 full 1 chrl: 156820106-156863034
KCNJ6 NM 002240 intronless 21 - chr21 :37917656-38220566
KCNMBl NM 004137 full 5 chr5: 169736744- 169759216
KCNMB2 NM 181361 intronless 3 + chr3 : 179726925- 180045918
KCNMB3 NM 014407 full 3 - chr3:180442257-180462105
KCNMB4 NM 014505 full 12 + chrl2:69036328-69112244
KCNNl NM 002248 full 19 + chrl9:17913110-17971929
KCNN2 NM 021614 intronless 5 + chr5: l 13715914-113861095
KCNN3 NM 002249 full 1 - chrl:151491989-151665827
KCNN4 NM 002250 full 19 - chrl9:48961526-48987249
KCNQ2 NM 172107 full 20 - chr20:61506985-61584437
KCNQ3 NM 004519 intronless 8 - chr8: 133209438-133572186
KCNQ4 NM 004700 full 1 + chrl:40908776-40974453
KCNQ5 NM 019842 intronless 6 + chr6:73378555-73963300
LGIl NM 005097 full 10 + chrl0:95497667-95548905
LGI2 NM 018176 full 4 - chr4:24678658-24718584
LGI3 NM 139278 full 8 - chr8:22059289-22080289
LGI4 NM 139284 full 19 - chrl9:40306256-40327944
LRRC7 NM 020794 func only 1 + chrl .-69927878-70300947
ME2 NM_002396 full 18 + chrl8:46649432-46729256
MVP NM 017458 func only 16 + chrl6:29729288-29767842
NAPA NM 003827 func only 19 - chrl9:52681702-52720309
NOVAl NM 002515 intronless 14 - chrl4:25983929-26146800
NRl 12 NM 022002 full 3 + chr3: 120974246-121021021
NSF NM 006178 func only 17 + chrl7:42013397-42190993
NT5E NM 002526 func only 6 + chr6:86206527-86263215
NTRK2 NM 006180 func only 9 + chr9:84503019-84869059
ODCl NM 002539 full 2 - chr2:10530105-10549051
PAFAHlBl NM 000430 full 17 + chrl7:2433685-2536638
PDYN NM 024411 full 20 chr20:1906402-1932702
PIK3CG NM 002649 func only 7 + chr7: 106089874-106142536
RAB3A NM 002866 func only 19 - chrl9:18167610-18185839
RAB3B NM 002867 func only 1 - chrl :52095857-52178369
RAB3C NM 138453 func only 5 + chr5:57904695-58184162
RAB3GAP NM 012233 func only 2 + chr2: 135633584-135762276 Gene name(a) ref seq(b) gene status'0' Chr(d> Strand(e Genomic address(f)
RALBPl NM 006788 func only 18 + chrl 8:9455529-9529105
RIMSl NM 014989 func only 6 + chr6:72643447-73168137
RPH3A NM 014954 func only 12 + chrl 2: 1 1 1682461-11 1798964
RYR3 NM 001036 func only 15 + chrl5:31380468-31946594
SCNlA NM 006920 full 2 - chr2 : 166672232-166765638
SCNlB NM 199037 full 19 + chrl9:40203373-40218014
SCN2A2 NM 021007 full 2 + chr2 : 165966098- 166073046
SCN2B NM 004588 full 11 - chrl l : 117537729-1 17562546
SCN3A NM 006922 full 2 chr2: 165768546- 165896060
SCN3B NM 018400 full 11 chrl 1 : 123006132-123040522
SCN4A NM 000334 full 17 chrl7:59368645-59414010
SCN5A NM 000335 full 3 - chr3:38563557-38676167
SCN8A NM 014191 intronless 12 + chrl2:50261286-50489365
SCN9A NM 002977 full 2 - chr2: 166879314-167003821
SLCl 2 A5 NM 020708 full 20 + chr20 :44081244-44123196
SLCl 7 A6 NM 020346 full 11 + chrl 1 :22306242-22358619
SLC17A7 NM 020309 full 19 - chrl9:54623469-54646596
SLCl 7 A8 NM 139319 full 12 + chrl2:99243461-99319304
SLClAl NM 004170 intronless 9 + chr9:4470443-4578469
SLC 1A2 NM 004171 intronless 11 chrl 1 :35228328-35407372
SLC 1A6 NM 005071 full 19 chrl9: 14920990-14954730
SLC 1A7 NM 006671 full 1 chrl :53263876-53330270
SLC28A1 NM 004213 func only 15 + chrl5:83218916-83291030
SLC29A3 NM 018344 func only 10 + chrl0:72739037-72794146
SLC32A1 NM 080552 full 20 + chr20:36776518-36792429
SLC6A1 NM 003042 full 3 + chr3: 10999455-1 1056934
SLC6A11 NM 014229 full 3 + chr3: 10822916-10956144
SLC6A12 NM 003044 full 12 chrl2: 168512-202753
SLC6A13 NM 016615 full 12 chrl2: 199051-252263
SLC6A2 NM 001043 func only 16 + chrl 6:54238056-54296199
SLCO 1A2 NM 134431 func only 12 - chrl2:21312093-21449638
SLCOlCl NM 017435 func only 12 + chrl2:20729665-20798585
SLCO2A1 NM 005630 func only 3 - chr3: 135133239-135241426
SLCO2B1 NM 007256 func only 11 + chrl 1 :74529810-74595945
SLCO3A1 NM 013272 func only 15 + chrl5:90187949-90508780
SLCO4A1 NM 016354 func only 20 + chr20:60734241-60775092
SLCO5A1 NM 030958 func only 8 - chr8:70746131-70919762
SLCO6A1 NM 173488 func only 5 - chr5: 101734552-101872619
SNAP25 NM 003081 func only 20 + chr20: 10137476-10237065
SSTRl NM 001049 func only 14 + chrl4:37736954-37753019
SSTR2 NM 001050 func only 17 + chrl 7:68662754-68680655
STXlO NM 003765 func only 19 - chrl9: 13115224-13131987
STX 12 NM 177424 func only 1 + chrl :27773924-27835956
STXlA NM 004603 func only 7 chr7:72557190-72588613
STX 1B2 NM 052874 func only 16 chrl 6:30909868-30939276
STX3A NM_004177 func only 11 + chrl 1 :59269464-59327752
STX5A NM 003164 func only 11 chrl 1 :62329945-62366136
STX6 NM 005819 func only 1 - chrl : 177672835-177733703
STX7 NM 003569 func only 6 chr6: 132821848-132885859
STX8 NM 004853 func only 17 chrl7:9093513-9430000
STXBPl NM 003165 func only 9 + chr9: 127444121-127535549
SV2A NM 014849 full 1 - chrl : 146687312-146712503
SV2B NM 014848 full 15 + chrl5:89560106-89640652
SV2C NM 014979 intronless 5 + chr5:75405060-75658172
SYNl NM 006950 func only X - chrX:47186554-47245510 Gene name(a) ref seq(b) gene status(c) Chr(d) Strand00 Genomic address'0
SYN2 NM 133625 func only 3 + chr3: 12010864-12208885
SYN3 NM 003490 fϊinc only 22 - chr22:31232094-31737237
SYNGRl NM 004711 func only 22 + chr22:38060453-38107079
SYNGR3 NM 004209 func only 16 + chrl6:1969968-1985276
SYTl NM 005639 func only 12 + chrl2:78103750-78348308
SYT2 NM 177402 func only 1 - chrl:199295544-199421202
TRAPPC4 NM_016146 func only 11 + chrl 1 :118384450-1 18400592
UGTl common na full 2 + chr2:234454803-234465802
UGT2B7 NM 001074 full 4 + chr4:70132984-70160464
VAMP2 NM 014232 func only 17 chrl7:8002188-8017017
VTIlA NM 145206 func only 10 + chrlθ: l 14187005-114488522
(a) Gene Name = Generic name allocated to that gene.
(b) Refseq = RefSeq Genes refers to known protein-coding genes taken from NCBI mRNA reference sequences collection. (c) Gene status = Full: gene considered for both functional and tagger SNP selection.
Func only: SNP selected for functional SNP selection only.
Intronless: Gene considered for both functional and tSNP selection but introns greater than 50kb not considered in tSNP selection, (d) Chr = Chromosome on which gene is located. (e) Strand = strand on which gene is located.
(f) Genomic address = The region considered for either or both of functional and tSNP selection.
Table 2A1 Dataset of SNPs and candidate genes
Figure imgf000017_0001
Table 2Aπ
Dataset of SNPs and candidate genes
Figure imgf000018_0001
Table 2B Dataset of SNPs and candidate genes
Figure imgf000018_0002
Table 2C Dataset of SNPs and candidate genes
Figure imgf000018_0003
Table 2D Dataset of SNPs and candidate genes
Figure imgf000019_0001
Table 2E Dataset of SNPs and candidate genes
Figure imgf000019_0002
Table 2F Dataset of SNPs and candidate genes
Figure imgf000019_0003
Table 2G Dataset of SNPs and candidate genes
Figure imgf000020_0001
Table 2H Dataset of SNPs and candidate genes
Figure imgf000021_0001
Table 21 Dataset of SNPs and candidate genes
Figure imgf000021_0002
DETAILED DESCRIPTION
[0036] The singular forms "a", "an" and "the" include plural aspects unless the context clearly dictates otherwise. Thus, for example, reference to "a polymorphism" includes a single nucleotide polymorphism (SNP) as well as two or more polymorphisms; reference to "an adverse drug reaction" or "an ADR" includes a single ADR, as well as two or more ADRs; reference to "the invention" includes a single aspect or multiple aspects of an invention; and so forth.
[0037] The present invention provides datasets of target genes (first knowledge base) or mutations in target genes wherein two or more mutations enable a correlation (second knowledge base) to be made with respect to the pharmacoresistance or pharmacosensitivity potential of a neurological medicament via an algorithmic training tool referred to as the hybrid classifier system. Candidate genes and/or SNPs are provided in Tables 2A (including Table 2A1 and 2A11) through 21 which represent training sets of data. Particularly, a mutation is screened for in two or more of GABBR2, KCNQl, SCN4B and SLCl A3. In a further embodiment, a mutation in one or more of KCNCl /MYODl, GRIA4 and GST A4 may be screened for in combination with a mutation in one or more of GABBR2, KCNQl, SCN4B and SLCIA3. Optimize training sets include two or more of the SNPs rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4). Hence, the present invention correlates genotype with predicted treatment outcomes. The correlation is via a bioinformatic analysis of genetic and clinical data in a pharmacogenomic approach to personalized medicine. It is proposed that individual genetic variation has multi-factorial implications in relation to drug absorption, distribution, metabolism, efflux, elimination and variability of drug target receptors which collectively or individually influence treatment outcomes. [0038] In yet a further embodiment, one or more of the above mutations may be detected in combination with a mutation in MDRl such as the SNP rslO45642.
[0039] Accordingly, a hybrid univariable/multivariate classification system is contemplated herein predictive of neurological disease treatment outcomes. Prediction rates of 70% or greater are provided herein using the hybrid classifier system. Clinicians can use this system to design personalized treatment programs for individual or cohort patients with reduced incidences of adverse drug reactions (ADRs) and/or poor responders. This increases the overall likelihood of treatment or prophylaxis success, improves health quality of the patient and reduces health costs. Reference to 70% or greater includes 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 and 100%.
[0040] The present invention further contemplates profiling or stratifying an individual or group of individuals with respect to therapeutic outcome potential in response to a particular drug or class of drugs in the treatment or prophylaxis of a neurological condition. Genotyping with respect to nucleotide mutations creates a genetic profile of a subject and this correlates to a likelihood of the subject responding favorably or not responding (e.g. having recurrence of symptoms or having an ADR) to a particular medicament.
[0041] By a "genetic profile" is meant that an individual or groups of individuals exhibiting a particular neurological condition which includes a neurological, psychiatric or psychological condition, phenotype or state or sub-threshold forms thereof or who are at the risk of developing same, exhibit two or more mutations at or within one or more genes selected from the list in Table 2 including its 5' or 3' terminal regions, promoter, exons or introns which is predictive of a therapeutic outcome. The genetic profile may be a single polymorphism (SNP) or mutli-nucleotide polymorphisms (MNPs) in a single gene or in a panel of genes, that is statistically significantly linked to a neurological condition. Reference to a "polymorphism" in this context includes a mutation. A mutation also includes a nucleotide insertion, addition, substitution and deletion as well as a rearrangement or microsatellite. Particular genes and/or mutations are provided in Tables 2A through 21. Most particularly, the SNPs contemplated herein comprise two or more of rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4). Optimized training sets of genes or mutation may differ with a particular population or sub-population including a geographical or ethnic sub-group of a population. One particular optimized training set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3). Another training set is rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4).
[0042] In yet a further embodiment, one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
[0043] Accordingly, the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2 which mutations correlate to potential responsiveness of the subject to the medicament.
[0044] A further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament. In this context, reference to a "positive" or "adverse" treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
[0045] Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition the method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2 which mutations correlate to potential responsiveness of the subject to the medicament.
[0046] The present invention further contemplates a method for identifying a genetic profile in a subject or group of subjects associated with the likelihood of a successful therapeutic outcome or otherwise to a neurological condition, the method comprising screening individuals for two or more polymorphisms including a mutation in a gene selected from the list in Table 2, including its 5' and 3' terminal regions, promoter, introns and exons which has a statistically significant linkage or association to a therapeutic outcome.
[0047] Reference to "Table 2" includes Tables 2A through 21 which represent training set of data. Table 2A includes Tables 2A1 and 2A11. It also includes particular SNPs such as optimized training sets selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4).
[0048] Accordingly, the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0049] A further aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0050] Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0051] Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0052] A further aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament. [0053] Still a another aspect of the present invention provides a method for a genotype- based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0054] A further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0055] Another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0056] Still a further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular neurological medicament, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0057] In yet a further embodiment, one or more of the above SNPs may be detected in combination with MDRl (rslO45642).
[0058] In this context, reference to a "positive" or "adverse" treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
[0059] The genetic locus comprising the genes listed in Table 2 may be referred to as the "gene", "nucleic acid", "locus", "genetic locus" or "polynucleotide". Each refers to polynucleotides, all of which are in the gene region including its 5' or 3' terminal regions, promoter, introns or exons. Accordingly, the genes of the present invention are intended to include coding sequences, intervening sequences and regulatory elements controlling transcription and/or translation. A genetic locus is intended to include all allelic variations of the DNA sequence on either or both chromosomes. Consequently, homozygous and heterozygous variations of the instant genetic loci are contemplated herein.
[0060] As indicated above, the present invention provides a genetic panel comprising different profiles of genes or mutations therein for different neurological conditions. Such profiles include polymorphisms, although any nucleotide substitution, addition, deletion or insertion or other mutation in one or more genetic loci is encompassed by the present invention when associated with a neurological condition. Accordingly, the present invention extends to rare mutations which although not present in larger numbers of individuals in a population, when the mutation is present in combination with at least one other mutation, it leads to a verifiable association between a responder or non-responder to a drug. The present invention is not to be limited to all the genes in the genetic panel but rather two or more genes in Table 2. Particular genes and mutations of interest are listed in Tables 2A1 and 2Aπ.
[0061] The term "polymorphism" or "mutation" refers to a difference in a DNA or RNA sequence or sequences among individuals, groups or populations which give rise to a statistically significant treatment outcome. Examples of genetic polymorphisms include mutations that result by chance, induced by external features or are inherited. [0062] Examples of nucleotide changes contemplated herein include single nucleotide polymorphisms (SNPs), multi-nucleotide polymorphisms (MNPs), frame shift mutations, including insertions and deletions (also called deletion insertion polymorphisms or DIPS), nucleotide substitutions, nonsense mutations, rearrangements and microsatellites. Two or more polymorphisms may also be used either at the same allele (i.e. haplotypes) or at different alleles. All these mutations are encompassed by the term "polymorphism".
[0063] Neurological conditions include, psychiatric and psychological conditions, phenotypes and states. Examples contemplated by the present invention include conditions related to dopamine pathway function and the function of associated neurotransmitters
GABA, glutamate, serotonin including but are not limited to epilepsy, addiction, dementia, anxiety disorders, bipolar disorder, schizophrenia, Tourette's syndrome, obsessive compulsive disorder (OCD), panic disorder, PTSD, phobias, acute stress disorder, adjustment disorder, agoraphobia without history of panic disorder, alcohol dependence
(alcoholism), amphetamine dependence, brief psychotic disorder, cannabis dependence, cocaine dependence, cyclothymic disorder, delirium, delusional disorder, dysthymic disorder, generalized anxiety disorder, hallucinogen dependence, major depressive disorder, nicotine dependence, opioid dependence, paranoid personality disorder, Parkinson's disease, schizoaffective disorder, schizoid personality disorder, schizophreniform disorder, schizotypal personality disorder, sedative dependence, shared psychotic disorder, smoking dependence and social phobia.
[0064] One particular example of a neurological condition is epilepsy or a related disorder. Two or more mutations in the genes in Table 2 can predict a treatment outcome for epilepsy or its related disorder. Particular examples of mutations are set forth in Tables 2B through 21. Most particularly, the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs391 1833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4).
[0065] Accordingly, the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0066] A further aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0067] Another aspect of the present invention contemplates a method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of epilepsy or related condition, the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0068] Another aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0069] A further aspect of the present invention provides a method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0070] Still yet another aspect of the present invention provides a method for a genotype- based prediction of responsiveness of a subject to a medicament in the treatment of epilepsy or a related condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0071] A further aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0072] Another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament. [0073] Still another aspect of the present invention is directed to a method of determining the potential or likelihood of a subject having a positive or adverse treatment response to a particular AED, the method comprising the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPs correlate to potential responsiveness of the subject to the medicament.
[0074] In this context, reference to a "positive" or "adverse" treatment response includes pharmacosensitivity, pharmacoresistance and/or an ADR to a particular neurological medicament.
[0075] The present invention further contemplates a method for identifying a genetic profile in a subject or group of subjects associated with the likelihood of a successful therapeutic outcome or otherwise to epilepsy or a related condition, the method comprising screening individuals for two or more polymorphisms including a mutation in a gene selected from the list in Table 2, including its 5' and 3' terminal regions, promoter, introns and exons which has a statistically significant linkage or association to a therapeutic outcome. Most particularly, the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GST A4).
[0076] As indicated above, particular genes or mutations are set forth in Tables 2A through 21. Most particularly, the SNPs contemplated herein comprise rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs 182623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs3911833 (KCNC1/MYOD1), rs507450 (GRIA4) and rsl82623 (GSTA4). Any selection of genes or mutations may be an optimized training set of data depending on the population or sub- population.
[0077] In yet a further embodiment, one or more of the above SNPs may be detected in combination with MDRl (rs 1045642).
[0078] Generally, the genetic test may be part of an overall diagnostic protocol involving clinical assessment and the diagnostic tools. Consequently, this aspect of the present invention may be considered as part of a therapeutic protocol.
[0079] Reference herein to an "individual" or a "subject" includes a human which may also be considered a patient, host, recipient or target. As indicated above, the present invention extends to veterinary applications.
[0080] The present invention enables, therefore, a stratification of individuals based on a genetic profile. The stratification or profiling enables a prediction of which treatment is likely to be most successful or appropriate or result in less recurrence or reduced adverse drug reaction.
[0081] There are many methods which may be used to detect a particular genotype. Direct DNA sequencing, either manual sequencing or automated fluorescent sequencing can detect sequence variation including a polymorphism or mutation. Another approach is the single-stranded conformation polymorphism assay (SSCP) [Orita, et al, Proc. Natl. Acad. ScL USA. 86:2166-2110, 1989]. This method does not detect all sequence changes, especially if the DNA fragment size is greater than 200 bp, but can be optimized to detect most DNA sequence variation. The reduced detection sensitivity is a disadvantage, but the increased throughput possible with SSCP makes it an attractive, viable alternative to direct sequencing for mutation detection. The fragments which have shifted mobility on SSCP gels are then sequenced to determine the exact nature of the DNA sequence variation. Other approaches based on the detection of mismatches between the two complementary DNA strands include clamped denaturing gel electrophoresis (CDGE) [Sheffield et al, Proc. Natl. Acad. Sci. USA 86:232-236, 1989], heteroduplex analysis (HA) [White et al, Genomics 72:301-306, 1992] and chemical mismatch cleavage (CMC) [Grompe et al, Proc. Natl. Acad. Sci. USA S<5:5855-5892, 1989]. None of the methods described above detects large deletions, duplications or insertions, nor will they detect a mutation in a regulatory region or a gene. Other methods which would detect these classes of mutations include a protein truncation assay or the asymmetric assay. A review of currently available methods of detecting DNA sequence variation can be found in Kwok (Curr Issues MoI. Biol. 5(2):43-60, 2003, Twyman and Primrose (Pharmacogenomics. 4(1):61-19, 2003), Edwards and Bartlett {Methods MoI. Biol. 226:287-294, 2003) and Brennan (Am. J. Pharmacogenomics. 1 (4):395-3Q2, 2001). Once a mutation is known, an allele-specific detection approach such as allele-specific oligonucleotide (ASO) hybridization can be utilized to rapidly screen large numbers of other samples for that same mutation. Such a technique can utilize probes which are labeled with gold nanoparticles or any other reporter molecule to yield a visual color result (Elghanian et al, Science 277:1078-1081, 1997).
[0082] A rapid preliminary analysis to detect polymorphisms in DNA sequences can be performed by looking at a series of Southern blots of DNA cut with one or more restriction enzymes, preferably with a large number of restriction enzymes. Each blot contains a series of normal individuals and a series of individuals having neurologic or neuropsychiatric diseases or disorders or any other neurological, psychiatric or psychological condition, phenotype or state. Southern blots displaying hybridizing fragments (differing in length from control DNA when probed with sequences near or to the genetic locus being tested) indicate a possible mutation or polymorphism. If restriction enzymes which produce very large restriction fragments are used, then pulsed field gel electrophoresis (PFGE) is employed. Alternatively, the desired region of the genetic locus being tested can be amplified, the resulting amplified products can be cut with a restriction enzyme and the size of fragments produced for the different polymorphisms can be determined. [0083] Detection of point mutations may be accomplished by molecular cloning of the target genes and sequencing the alleles using techniques well known in the art. Also, the gene or portions of the gene may be amplified, e.g., by PCR or other amplification technique, and the amplified gene or amplified portions of the gene may be sequenced.
[0084] Additionally, real-time PCR such as the allele specific kinetic real-time PCR assay can be used or allele specific real-time TaqMan probes.
[0085] For allele-specific PCR, primers are used which hybridize at their 3' ends to a particular target genetic locus or mutation. If the particular polymorphism or mutation is not present, an amplification product is not observed. Amplification Refractory Mutation System (ARMS) can also be used, as disclosed in European Patent Application Publication No. 0332435. Insertions and deletions of genes can also be detected by cloning, sequencing and amplification. In addition, restriction fragment length polymorphism (RFLP) probes for the gene or surrounding marker genes can be used to score alteration of an allele or an insertion in a polymorphic fragment. Such a method is particularly useful for screening relatives of an affected individual for the presence of the mutation found in that individual. Other techniques for detecting insertions and deletions as known in the art can be used.
[0086] In SSCP, DGGE and the RNase protection assay, an electrophoretic band appears which is absent if the polymorphism or mutation is not present. SSCP detects a band which migrates differentially because the sequence change causes a difference in single-strand, intramolecular base pairing. RNase protection involves cleavage of the mutant polynucleotide into two or more smaller fragments. DGGE detects differences in migration rates of mutant sequences compared to wild-type sequences, using a denaturing gradient gel. In an allele-specific oligonucleotide assay, an oligonucleotide is designed which detects a specific sequence, and the assay is performed by detecting the presence or absence of a hybridization signal, In the mutS assay, the protein binds only to sequences that contain a nucleotide mismatch in a heteroduplex between mutant and wild-type sequences. [0087J Mismatches, according to the present invention, are hybridized nucleic acid duplexes in which the two strands are not 100% complementary. Lack of total homology may be due to deletions, insertions, inversions or substitutions. Mismatch detection can be used to detect point mutations in the gene or in its mRNA product. While these techniques are less sensitive than sequencing, they are simpler to perform on a large number of samples. An example of a mismatch cleavage technique is the RNase protection method. In the practice of the present invention, the method involves the use of a labeled riboprobe which is complementary to the human wild-type genes (i.e. such as those listed in Table 2). The riboprobe and either mRNA or DNA isolated from the person are annealed (hybridized) together and subsequently digested with the enzyme RNase A which is able to detect some mismatches in a duplex RNA structure. If a mismatch is detected by RNase A, it cleaves at the site of the mismatch. Thus, when the annealed RNA preparation is separated on an electrophoretic gel matrix, if a mismatch has been detected and cleaved by RNase A, an RNA product will be seen which is smaller than the full length duplex RNA for the riboprobe and the mRNA or DNA. The riboprobe need not be the full length of the mRNA or gene but can be a segment of either. If the riboprobe comprises only a segment of the mRNA or gene, it will be desirable to use a number of these probes to screen the whole mRNA sequence for mismatches.
[0088] hi similar fashion, DNA probes can be used to detect mismatches, through enzymatic or chemical cleavage (see, for example, Cotton et al, Proc. Natl. Acad. Sci. USA 57:4033-4037, 1988; Shenk et al, Proc. Natl. Acad. Sd. USA 72:989-993, 1975; Novack et al, Proc. Natl. Acad. Sd. USA 55:586-590, 1986). Alternatively, mismatches can be detected by shifts in the electrophoretic mobility of mismatched duplexes relative to matched duplexes (see, for example, Cariello Am. J. Human Genetics 42:726-734, 1988). With either riboprobes or DNA probes, the cellular mRNA or DNA which might contain a mutation can be amplified using PCR (see below) before hybridization. Changes in DNA of the associated genetic polymorphisms or genetic loci can also be detected using Southern blot hybridization, especially if the changes are gross rearrangements, such as deletions and insertions. [0089] Once the site containing the polymorphisms has been amplified, the SNPs can also be detected by primer extension. Here a primer is annealed immediately adjacent to the variant site, and the 5' end is extended a single base pair by incubation with di- deoxytrinucleotides. Whether the extended base was a A, T, G or C can then be determined by mass spectrometry (MALDI-TOF) or fluorescent flow cytometric analysis (Taylor et al, Biotechniques 30:661-669, 2001) or other techniques.
[0090] Nucleic acid analysis via microchip technology is also applicable to the present invention. In this technique, thousands of distinct oligonucleotide probes are built up in an array on a silicon chip. Nucleic acids to be analyzed are fluorescently labeled and hybridized to the probes on the chip. It is also possible to study nucleic acid-protein interactions using these nucleic acid microchips. Using this technique, one can determine the presence of mutations or even sequence the nucleic acid being analyzed or one can measure expression levels of a gene of interest. The method is one of parallel processing of many, including thousands, of probes at once and can tremendously increase the rate of analysis.
[0091] Mutations falling outside the coding region of the target loci can be detected by examining the non-coding regions, such as introns and regulatory sequences near or within the genes. An early indication that mutations in non-coding regions are important may come from
[0092] Alteration of mRNA expression from the genetic loci can be detected by any techniques known in the art. These include Northern blot analysis, PCR amplification and
RNase protection. Diminished mRNA expression indicates an alteration of the wild-type gene. Alteration of wild-type genes can also be detected by screening for alteration of wild-type protein. For example, monoclonal antibodies immunoreactive with a target protein (i.e. two or more proteins encoded by one or more genes listed in Table 2) can be used to screen a tissue. Lack of cognate antigen or a reduction in the levels of antigen would indicate a mutation. Antibodies specific for products of mutant alleles could also be used to detect mutant gene product. Such immunological assays can be done in any convenient formats known in the art. These include Western blots, immunohistochemical assays and ELISA assays. Any means for detecting an altered protein can be used to detect alteration of the wild-type protein. Functional assays, such as protein binding determinations, can be used. In addition, assays can be used which detect the protein biochemical function. Finding a mutant gene product indicates alteration of a wild-type gene product.
[0093] Hence, the present invention further extends to a method for identifying a genetic basis behind a successful or adverse treatment protocol for a neurological condition in an individual, the method comprising obtaining a biological sample from the individual and detecting two or more mutations in one or more proteins encoded by one or more genes listed in Table 2.
[0094] In an embodiment, the neurological condition is epilepsy or a related condition.
[0095] The altered amino acid sequence may be detected via specific antibodies which can discriminate between the presence or absence of an amino acid change, by amino acid sequencing, by a change in protein activity or cell phenotype and/or via the presence of particular metabolites if the protein is associated with a biochemical pathway.
[0096] A mutant gene or corresponding gene products can also be detected in other human body samples which contain DNA, such as serum, stool, urine and sputum. The same techniques discussed above for detection of mutant genes or gene products in tissues can be applied to other body samples. By screening such body samples, an early determination can be achieved for subjects on a particular drug or about to be prescribed a particular drug.
[0097] The present invention extends to two or more isolated oligonucleotides which comprise from about three to about 1000 consecutive nucleotides from the gene or its corresponding cDNA or mRNA as listed in Table 2 which encompass at least two polymorphisms or mutations associated with a particular therapeutic outcome for a neurological condition. In an embodiment, the neurological condition is epilepsy or a related condition.
[0098] In one embodiment, one of the at least two primers is involved in an amplification reaction to amplify a target sequence. If this primer is also labeled with a reporter molecule, the amplification reaction will result in the incorporation of any of the label into the amplified product. The terms "amplification product" and "amplicon" may be used interchangeably.
[0099] The primers and the amplicons of the present invention may also be modified in a manner which provides either a detectable signal or aids in the purification of the amplified product.
[0100] A range of labels providing a detectable signal may be employed. The label may be associated with a primer or amplicon or it may be attached to an intermediate which subsequently binds to the primer or amplicon. The label may be selected from a group including a chromogen, a catalyst, an enzyme, a fluorophore, a luminescent molecule, a chemiluminescent molecule, a lanthanide ion such as Europium (Eu34), a radioisotope and a direct visual label. In the case of a direct visual label, use may be made of a colloidal metallic or non-metallic particular, a dye particle, an enzyme or a substrate, an organic polymer, a latex particle, a liposome, or other vesicle containing a signal producing substance and the like. A large number of enzymes suitable for use as labels is disclosed in U.S. Patent Nos. 4,366,241, 4,843,000 and 4,849,338. Suitable enzyme labels useful in the present invention include alkaline phosphatase, horseradish peroxidase, luciferase, β- galactosidase, glucose oxidase, lysozyme, malate dehydrogenase and the like. The enzyme label may be used alone or in combination with a second enzyme which is in solution. Alternatively, a fluorophore which may be used as a suitable label in accordance with the present invention includes, but is not limited to, fluorescein-isothiocyanate (FITC), and the fluorochrome is selected from FITC, cyanine-2, Cyanine-3, Cyanine-3.5, Cyanine-5, Cyanine-7, fluorescein, Texas red, rhodamine, lissamine and phycoerythrin. [0101] In order to aid in the purification of an amplicon, the primers or amplicons may additionally be incorporated on a bead or other solid support.
[0102] In order to detect the presence of alleles from the genes from Table 2 collated to a particular therapeutic outcome for a neurological condition such as epilepsy, a biological sample such as blood is obtained and analyzed for the presence or absence of a panel of target alleles comprising from about two to 100 alleles or from about two to 50 alleles or from two to about 30 alleles of the genetic loci identified as being statistically significantly associated with the treatment outcome for epilepsy. Results of these tests and interpretive information are returned to the health care provider a decision on which medicament is appropriate. Such diagnoses may be performed by diagnostic laboratories, or, alternatively, diagnostic kits are manufactured and sold to health care providers or to private individuals for self-diagnosis. Suitable diagnostic techniques include those described herein as well as those described in US Patent Numbers 5,837,492; 5,800,998 and 5,891,628.
[0103] The present invention is now described by reference to the following non-limiting Examples. A range of methods is employed in these Examples as described below.
General Methods
Cohort Recruitment and Blood Collection
[0104] 237 newly treated epilepsy patients were prospectively recruited from epilepsy clinics Australia- wide. Individuals were followed up at three periodical intervals: 3 months, 1 year and 2 years after baseline. At baseline, the initial AED was prescribed at the discretion of the treating neurologist according to epilepsy type and syndrome. Once consent was provided, 2 x 9mL blood samples were obtained via phlebotomy from a vein on the back of the hand or just below the elbow of the patient and these samples were frozen at 2-60C for genotyping purposes.
DNA Purification
[0105] Purification of DNA from whole blood was performed using the QIAamp Blood Maxi Kit (Spin Protocol); this protocol is for purification of genomic DNA from up to 10ml of whole blood. In brief: cells are lysed using a protease, this reaction is then stopped and the lysate is loaded onto a QIAamp Maxi column. A series of purification steps follow before the final centrifugation of the eluate containing the DNA is performed as described herein after.
1. Pipet 500μl QIAGEN Protease into 50ml centrifuge tube.
2. Add 5-1OmI blood and mix briefly.
3. Add 12ml Buffer AL, and mix thoroughly by inverting tube 15 times, followed by vortexing for 1 minute. To ensure sufficient lysis, samples must be thoroughly mixed to yield a homogenous solution.
4. Incubate at 7O0C for approximately 15 minutes.
5. Add 10ml ethanol to the sample, and mix by inverting 10 times, followed by additional vortexing. 6. Carefully transfer half of the solution from previous step onto the QIAamp Maxi column placed in a 50ml centrifuge tube. Close the cap and centrifuge at 3000rpm for 3 minutes.
7. Remove the QIAamp Maxi column, discard the filtrate from the centrifuge tube, and place the QIAamp Maxi column back into the 50ml centrifuge tube. Load the remainder of the solution from step 5 onto the QIAamp Maxi column. Close the cap and centrifuge again at 3000rpm for 3 minutes.
8. Remove the QIAamp Maxi column, discard the filtrate from the centrifuge tube, and place the QIAamp Maxi column back into the 50ml centrifuge tube.
9. Add 5ml Buffer AWl to the QIAamp Maxi column. Close the cap and centrifuge at 5000rpm for 1 minute.
10. Add 5ml Buffer AW2 to the QIAamp Maxi column. Close the cap and centrifuge at 5000rpm for 15 minutes.
• If centrifugal force is below 4000 x g, incubate QIAamp Maxi column for 10 minutes at 7O0C in an incubator to evaporate residual ethanol.
11. Place QIAamp Maxi column in a clean 50ml centrifuge tube, and discard the collection tube containing the filtrate.
12. Pipet ImI Buffer AE/water, equilibrated to room temperature, directly onto membrane of the QIAamp Maxi column and close the cap. Incubate at room temperature for 5 minutes, and centrifuge at 5000rpm for 2 minutes.
• For long term storage, elute in Buffer AE and store in aliquots at -2O0C, to avoid acid hydrolysis of DNA if dissolved in water. 13. For maximum concentration: Reload the eluate containing the DNA (ImI) onto the membrane of the QIAamp Maxi column. Close the cap and incubate at room temperature for 5 minutes. Centrifuge at 5000rpm for 5 minutes.
• After maximum yield: Pipet ImI fresh Buffer AE onto the membrane of the QIAamp Maxi column. Close cap, incubate at room temperature for 5 minutes. Then centrifuge at 5000rpm for 5 minutes.
Study Design: Candidate gene selection
[0106] In work leading to the present invention, a candidate gene approach was adopted whereby the genes tested were selected based on prior biological knowledge of epilepsy and drug metabolism. This included all the known members of the voltage-gated sodium and calcium channels, a subset of members of the chloride and potassium channels, key receptors, metabolizers and transporters of the major neurotransmitters GABA, acetylcholine and glutamate and all major targets, transporters and metabolizers of the AEDs (Gianpiero and Weale, Lancet Neurology 6, 2007).
[0107] Due to the non-familial nature of the study, the candidate gene approach is ideal as unrelated cases and controls suffice. Although the whole genome approach is perhaps a better model to base the study on, it has financial limitation, and would cost a substantial amount more to organize and employ. Additionally, the candidate gene approach is appropriate as there is a significant understanding of the pathophysiological of epilepsy. A common concern with this approach is the lack in knowledge of functional variants, and this is why the study incorporates tagger SNPs that have no current functional significance.
These SNPs may not the casual variants, but may lie within or near the casual variant by virtue of linkage disequilibrium. Table 2 lists the candidate genes used in this study.
SNP selection strategy [0108] All common variation across the list of candidate genes in Table 2 were examined through a combined direct (genotyped) and indirect (captured via linkage disequilibrium) mapping strategy. Functional variation was assessed directly whilst all other variation of unknown function was assessed indirectly through tagger SNPs. Additionally, 100 SNPs were incorporated for stratification analysis; these SNPs are neutral with respect to epilepsy (based on current available literature) and spaced far apart so as to be in linkage equilibrium. A detailed description of the protocol used to select candidate genes, and furthermore, SNPs to be used is provided below.
[0109] SNP selection was by a fuse a tagging (map based) with a functional (sequence based) approach to detect variation functional in the development and treatment of common epilepsy treatment outcomes; using a candidate gene approach with selection of genes based on biology.
1. Run each gene through Tamal, a software package designed to publicly search available databases for variation that falls under one of the following categories: coding (located in exon), protein altering (non-synonymous or splice variants), promoter variant, conserved region variation (conserved at the 99th percentile across five species [human, chimp, mouse, rat, chicken]), variation with predicted regulatory potential or variation that affects transcription factors binding sites.
2. Variation that falls into one of the above categories is divided into two subsets - those geno typed by HapMap, and those that have not been genotyped by HapMap. HapMap functional are forced as tagging SNPs. Nevertheless functional variations not genotyped by HapMap are also included, for any analysis, functional and tags were examined thus covering both a map and sequence approach.
3. Stratification is always a concern with association studies. Although majority of the subjects are Caucasian, 100 neutral variants were genotyped to act as 'genomic controls'. These variants were selected at random from HapMap and are all located at least 50kb from the nearest known gene. This dataset allows the determination of a empirical estimate stratification and correct association results accordingly. Genotyping and Data Cleaning
[0110] DNA was extracted and purified from the blood samples of the first 179 patients who had at least one year follow-up and a technically adequate blood sample available. This DNA was genotyped for 4,041 SNPs from a pre-selected set of 279 candidate genes. Noise reduction was performed by removing SNPs with missing values or little variation amongst treatment outcomes; improving the quality of the available dataset.
Collection ofphenotypic data
[0111] Upon enrolment to the study, information including demographic, familial, imaging, cognitive, and physical characteristic data were collected. The patients ethnic background is acquired, their seizure type and frequency is recorded, details of AED prescribed and any other medications is collected, followed by questions regarding their height and weight.
[0112] At three months, one year and two year intervals, patients were contacted either whilst at the epilepsy clinic or via phone to complete follow-up information. The information regarded seizure control, and is an opportunity to discover whether the patient has experienced side effects, most ADRs tend to present themselves within the first few months of administration. Additional to these, the information was gathered on any imaging or EEg recordings that patients may have done since their last follow-up, medication changes are recorded, and a follow-up on their weight is noted.
Phenotyping
[0113] Patients were followed up upon commencement of AED and asked a series of questions. These questions formed the basis from which phenotypic diagnoses would be made. Non-responsiveness was defined as spontaneous seizures to initial drug therapy where the patient had reached therapeutic levels. In instances where a patient was taken off an initial AED before pharmacoresponsiveness could be determined by ADRs, pharmacoresponsiveness was determined on a subsequent AED used.
[0114] ADRs were phenotyped by a clinical neurologist. Reported ADRs were then clustered into five clinically relevant groups: neurological, metabolic, immunological, gastrointestinal (GIT), and other. This clustering provides an opportunity for deeper analysis by increasing the number of events within each general ADR group. Table 3 displays all patients from the genotyped cohort who experienced an ADR, which ADR, and the general class of ADRs it falls into.
Table 3 ADRs
Figure imgf000046_0001
Figure imgf000047_0001
Neurological has three sub-groups: (F&S) = Fatigue, Sedation, Ataxia and Depression
(M&N) = memory/concentration Difficulties and
Neurocognitive Tremor Metabolic has two sub-groups: GOW = Gain of Weight; >5kg within first 3 months OR >7.5kg over 2 year course
LOW = Loss of Weight; >5kg within first 3 months OR >7.5kg over 2 year course
Immunological has two sub-groups: Rash; subcutaneous allergic reactions believed to be drug related
Hepatitis; drug-induced
GIT has two sub-groups: Nausea and Diarrhoea Other includes: Hair loss and Acne Statistical Methods
[0115] The samples were analyzed by four different approaches: two univeriate (Fishers t- tests and chi-square, and using SAS Genetic Marker) and two multivariate (GeneRave and Hybrid approach). The analyses were completed on two treatment outcomes, responsiveness and ADRs. Once as a combined cohort then further subdivided based on the two most commonly prescribed AEDs carbamazepine and valproate.
Univariate single SNP analysis [0116] For each individual at any given SNP there are two possible values "Allele A" or "Allele B", making three possible genotypes "AA", "AB" or "BB". Each SNP was analysed independently at both the allelic and genotypic level between the two treatment outcome classes for each of the two treatment outcomes being analysed. A number of statistics were outputted in Microsoft Ecel: p-value, chi-square test, allelic and genotypic odds ratios (OR) the latter limited for the comparison between homozygous genotypes. All statistics were coded within Microsoft Excel.
[0117] Additionally, 100 randomized control runs were made, each run having the same number of cases vs controls as the pharmacoresponsiveness findings, p-values for the top 10 SNPs were noted providing an empirical insight into how significant the treatment outcome findings for pharmacoresistance were. This relatively weak empirical control is preferred over the over-conservative Bonferroni correction as many of the SNPs are related and thus inherited together.
SAS Genetic Marker
[0118] A combination of PROC CASECONTROL, and PSMOOTH were adopted to locate SNPs that may affect susceptibility to AED treatment outcome.
[0119] In the CASECONTROL procedure, SNP data from unrelated individuals were divided into two classes, i.e. those which represent the non-responders (Case) and those representing the responders (Control). The genotype and allele case-control chi-square tests along with the linear trend test were chosen and applied to test for both dominant and additive allele effects on treatment outcome penetrance. This analysis is to determine the contributions of individual SNPs. Therefore, the trend test was modified to take into account collinearity by applying the Variance Inflation Factor (VIF) [Devlin and Roeder, Biometrics 55:997-1004, 1999] to the trend chi-square test as a genomic control. Collinearity accounts for the SNP redundancy observed when some SNPs are predicted by others.
[0120] The PSMOOTH procedure was applied to the data used in CASECONTROL to reduce the number of false positives. This smoothing takes into consideration p-values from neighbouring and possibly correlated markers. To correct for the number of hypotheses, adjustments for multiple testing were made, i.e. both the Bonferroni,
Publicazioni del r Instituto Superiore di Science Economiche e Commerciali di irenze 8:3-
62, 1936 and Sidak, Journal of the American Statistical Association 62:626-633, 1967 multiple hypothesis tests, where Sidak test is slightly less conservative than the Bonferroni test.
[0121] The PROC %TPLOT function is a useful application in detecting blocks of LD and is methodologically similar to Haploview. It combines output from the CASECONTROL, PSMOOTH, and HAPLOTYPE procedures. It allows the visualization of smoothed p- values from Hardy Weinberg Equilibrium (HWE) tests, tests for linkage disequilibrium between SNPs and association tests between SNPs and pharmacoresi stance.
GeneRave [0122] GeneRave's RChip is an example of the many commercially available suites. The RChip suite was selected for its ability to analyse the relationship between response variables and a set of predictors when the number of predictors far exceeds the number of observations. This suite was designed based on the leukaemia data published by Yeoh et al, Cancer CU 7:133-143, 2002. RChip has the ability to identify genes/SNPs that discriminate between different phenotypes and this capability was tested using the epilepsy treatment outcome data. [0123] By approaching the problem with supervised learning, a predictor can be developed that classifies 'test data', and in doing so determines the features (SNPs) responsible for the classification. The aim being to find SNPs that based on presence can discriminate between two categorical classes. To facilitate discrimination, a scoring system is required to reflect the SNPs influence on the distinction between classes. The score adopted in this analysis is the signal-to-noise statistic (golub score), described more fully below.
[0124] The classifier is built in 2 steps: First, the most differentially expressed SNPs between the classes are selected using the golub score. Second, the £NN classifier evaluates the number of signature SNPs (N) and the number of nearest neighbours (k) to use by optimising the classification performance on validation datasets. The pseudo-code for the classifier program which I built using PERL programming language is provided herein under:
• Computer reads in training targets (classified treatment outcomes) and stores them in array @inarray.
• Computer reads in validation targets (classified treatment outcomes) and stores them in array @validarray. • Computer then feeds the training targets into the golub function and the results are returned into the ©complete array. — GOLUB FUNCTION —
• Computer reads in the training data and stores it into temporary array, @inarray.
• For each SNP in the training data, split them up into two arrays one for patients with treatment outcome 1, and the other for treatment outcome 2 based on training target file.
• For each SNP in the resulting treatment outcome 1 array add the value of the SNP for each patient to a scalar variable $meanθ
• For each SNP in the resulting treatment outcome 2 array add the value of the SNP for each patient to a scalar variable Smeanl • Divide the final accumulative mean for each SNP $meanθ by the number of patients in the training data who had treatment outcome 1.
• Divide the final accumulative mean for each SNP $meanl by the number of patients in the training data who had treatment outcome 2. • Calculate the standard deviation for treatment outcome 1 group.
• Calculate the standard deviation for treatment outcome 2 group.
• Calculate golub score to equal absolute difference in means divided by the difference in standard deviations between both classes for each SNP and store in array @golub.
— END — • Sort the returned golub array in ascending order and store it as a new array @dataS.
• Computer prompts the researcher to set the N value, $limit, specifying the number of SNPs to use in the N-dimensional space.
• For each of the N values chosen, create an array of arrays called ©universal.
• Computer then creates a hash reference that reads in the training data and then another hash reference for the validation data.
• Computer prompts researcher to enter k value, number of nearest neighbors to use in classification.
• Computer then stores the integer value for the number of entries in the validation set.
• Computer then stores the integer value for the number of entries in the training set. • The computer then takes each validation entry and checks it by the corresponding training entry to calculate the Euclidean distance between the training and the validation, and then pushes this into the array @calc.
• @calc now holds the comparison for each training set with the validation set, and sorts these in ascending order. • For the specified number of nearest neighbors (k), the computer loops through @calc and counts the number of treatment outcome 1 occurrences, and the number of treatment outcome 2 occurrences from the training data for that one validation kernel.
• Computer compares number of values in treatment outcome 1 to treatment outcome 2, and sets the variable Stally to equal the majority class of the two treatment outcomes. • Computer then has a counter which increments variable $count by one for every correct classification.
• Computer then prints out the number of correct classifications out of the number of validation tests.
[0125] The £NN classifier requires that there be non-null numbers for any of the three genotypes in any SNP. This includes when sub-dividing cohort into 5 sub-cohorts. Therefore SNPs that had multiple missing values or patients with multiple missing SNPs were removed from the analysis.
Gene Ranking - Golub Score
[0126] Based upon averages (x,) and standard deviations (σ,) of a SNPs presence in the two classes, for each SNP i the score w, is computed.
Figure imgf000052_0001
[0127] Where gene Vs average expressioiflevel, x,(m), in class m is calculated by summing over all Nm experiments in class m,
X1 H = I £ ,
"XT ^ x"κ "m kCm
[0128] And the standard deviation is,
Figure imgf000052_0002
[0129] Once all w, are calculated, the program generates a list of SNPs sorted by decreasing score is generated.
kNN classification [0130] Introduced by Fix and Hodges, Discriminatory analysis - non-parametric discrimination: Consistency properties (No. 4) Randolph Field, Taxas; USAF School of Aviation Medicine, 1951, &NN has been extensively studied within the fields of bioinformatics as a tool for classification and clustering in areas of pattern recognition, with the underlying principle being: similar objects belong to similar classes. Earlier studies such as: Crimins et al, In Methods of Microarray Data Analysis (Springer US): \9\- 205, 2005 - Lung cancer; Binder et al, Clinical and diagnostic laboratory immunology 72:1353-1357, 2005 - immunoassay based anti-nuclear antibody tests; and Kim et al, BMC bioinformatics 5:160, 2004 - microarray experiments, have shown that the &NN algorithm can obtain accurate and reliable results within medical fields.
[0131] The crux of the &NN classifier is that the class of a validation sample is decided by the majority class among its k nearest neighbours. A neighbour is deemed nearest if it has the smallest Euclidean distance in the ^-dimensional space, where N is the number of top ranked SNPs chosen.
[0132] First, the algorithm calculates the distance from validation sample y to each training sample x, using Euclidean distance. Then the researcher identifies the k training samples with the smallest distance to y, and checks what the majority class is amongst them. This is re-done for each validation sample. To avoid a tied vote, k is chosen as an odd number.
[0133] The &NN algorithm uses all attributes in the training set and plots them into the data space. When a new validation case is presented for classification, a kernel is formed in data space centred on the validation case. This kernel is hypersphere shaped and is just large enough to contain the k nearest neighbours of the validation case using a Euclidean distance metric. Classification is then performed according to the k nearest neighbours found in the kernel.
[0134] By modifying the parameters k and N many classifiers were created. A balance needs to be made when selecting k. Computing time increases as k increases and additionally if k is too large it would encompass the entire population; however advantage of higher k values is that it provides smoothing, thus reducing vulnerability to noise in the training data.
Validation Strategy [0135] Arguably the most important success criterion of this approach is the evaluation of the classifiers performance. Two strategies were trialled, the sturdy Training→ Validation→ Test method, which is referred to as the TVT method, and the cross-validation method. Initially the TVT method was applied; however this method consumed many cases in the validation and test sets, leaving the training set, which is the classification building step with a low number of cases to build the classifier upon. Due to this size limitation, the alternative cross-validation approach was adopted. The advantage is that all the data are used for cross-training and testing, whilst the validation remains completely independent to the training.
[0136] As it is difficult to obtain another large sample of classified cases to act as validation and testing sets; hi cross-validation, the single large set is used to both build the classifier and estimate classification accuracy. The original set L is randomly divided into V subsets, Lj ,L2,..., Ly, of as near equal size as possible and with equal number of responders and non-responders in each group. For every v, v=\,2,...,V a classifier is built using the combinatory set of (L- tv) as the training set used to classify cases in Lv. A test sample estimate is obtained for each v using:
NK
*R'S (d(v)) = 1/Nv Σ
Figure imgf000054_0001
= C1-) i=l '
[0137] Where (x,, c,) C Ly and Nv is the number of cases in L1/. The F-fold cross- validation estimate Rcv(d) is the average of the V test sample estimates:
V *RCV (d(v)) = XIV ∑ R's (ά(v)) v=l
*Calculations obtained from Nakhaeizadeh and Taylor, Machine Learning and Statistics: The Interface (John Wiley & Sons, Inc):27-35, 1997
[0138] V was chosen to be five cohorts in this analysis. Reasoning for five cohorts: getting the optimal training (classification building) dataset as more individuals would be used in building the classifier. Thus, by dividing the cohort of 119 patients into five groups it allows for a training dataset of approximately 100 patients on which to build the classifier and on average 23 in the validation cohorts. Ideally larger validation sets would cater for any noise that may affect the classification, however due to dataset limitations, 23 individuals in the validation set was the target number taking all aspects into consideration.
[0139] As mentioned, with relatively small validation sets, noise in the form of false positive would always be an issue. Therefore, the top 15 ranked SNPs for each of the V test sample estimates were also recorded. These SNPs were used as a cross-check to determine which five SNPs reappear in the top 15 across the V test sample estimates. Once the top five SNPs were selected, classification was re-run for each validation cohort using the combined top five SNPs. Fishers t-tests were then performed for each validation cohort to determine the strength of the classifier for each.
Replication
[0140] Following analysis, a further 49 'newly treated' epilepsy patients reached 12-month follow-up maturity. These patients were then prospectively genotyped exclusively on the top 5 SNPs used by the classifier and an additional 3 SNPs which were of interest at the univariate level.
[0141] The 49 patients are independent of the initial 179 patients that were genotyped. They were passed through the classifier and also SAS Genetics Marker code. Regardless of model derivation, a significant result in this replication would suffice to prove the effect that the 5 SNPs used in the classifier have in predicting pharmacoresponsiveness. [0142] In addition to this, the classifier and SNPs were tested on a population of community recruited 189 chronic treated epilepsy patients. Initial phenotype data included patients without a follow up (21), patients without a recorded AED (22), patients without recorded seizure (15), patients with date errors (23), were removed leaving final analysis of 108 patients. These 108 patients were used to validate classifiers performance in a separate population from which the classifier was built, and the 8 SNPs were looked at individually.
[0143] The original model was developed with the flexibility of including two-year follow-ups when deciding responsiveness, the chronic cohort required a much stricter 12 month follow-up restriction. Previous studies of pharmacoresistance in chronic epilepsy populations have set the responder criterion to less than 3 seizures within the year. Flexibility is allowed for in the restrictions to patients who have been seizure free within 12 months of their interview. 12 months of seizure freedom is required for obtaining a motor-vehicle license in Australia.
EXAMPLE 1
Cohort characteristics
[0144] At the time of initial genotyping, 237 patients were recruited into the pharmacogenetics and epilepsy study. Of these, 179 patients had available DNA which was genotyped. Selection for genotyping was based unbiasedly upon patients with extracted DNA. There was no statistical difference between those genotyped and those not genotyped (Table 4). Of the 179 patients, five withdrew (3%), eight stopped taking AED (4%), two were believed to be having pseudoseizures (1%), 12 lost contact *7%), 15 patients had no follow-up clinical data available (4%), 23 had vast amounts of missing SNP genotypes (13%) and six had a smaller number of missing SNP genotypes (3%). This left 121 individuals available for univariate analysis (68%) and 115 for the classifier (64%).
Table 4 Cohort characteristics
Figure imgf000057_0001
EXAMPLE 2 Anti-epileptic drug retention
[0145] Pharmacoresistance and ADRs are the two main reasons why patients cease treatment with an AED. A significant proportion of patients in the cohort had ceased or changed the initial AED prescribed, with an overall retention rate of only 59% by the end of the two year follow-up period. The retention rate was similar for the two most commonly prescribed AEDs in this cohort; 57% for patients prescribed carbamazepine
(CBZ) and 63% for patients prescribed valproate (VPA) over the two-year follow up (Table 5),
Table 5 Retention rate of the AEDs CBZ and VPA in this study
Figure imgf000058_0001
EXAMPLE 3 Seizure Control
[0146] Of the patients who met the 12 month follow-up criteria; at 3 months, 17% experienced spontaneous seizures (not associated with medication non-compliance or extreme provoking circumstances), at one year a further 6% experienced spontaneous seizures and by the two year period only 72% of newly treated patients remained spontaneous seizure free (Table 6).
Table 6 Seizure control amongst genotyped patients over the two year period
Figure imgf000059_0001
EXAMPLE 4 Univariate SNP statistics
[0147] 121 patients with 12 months follow-up were analyzed. The top ranked SNP was from the GABBR2 also known as the GPR51 gene (rs2808526) with a p-value of 0.0006 and a genotypic OR of 21 (Table 7).
Table 7 Pharmacoresponsiveness analysis
Figure imgf000060_0001
[0148] A program the re-run univariate analysis, randomising patients as responders or non-responders for 100 control runs, the average p-value obtained was 0.005, with only four p-values occurring lower than the actual 0.0006 p-value found for the GABBR2 also known as GPR51 SNP (Table 8). Table 8 Empirical control
Figure imgf000061_0001
a Average p-value is the average value of the 100 control runs b The real result indicates the actual top p-value obtained for the GABBR2 rs2808526 SNP c The number of 100 control runs that obtained a p-value greater than the actual result.
[0149] CBZ specific analysis (n=72) resulted in no highly significant SNPs standing out. The top SNP was from the GLUL gene with a p-value of 0.07 and an OR of 9.2. This decrease in significance implies it has an effect but is unlikely to be drug specific. VPA specific analysis (n=45) resulted in a SNP from the CLCN7 gene standing out from the rest with a p-value of 0.0006 and a very high OR of 35. However, as expected, this high significance did not remain after Bonferroni correction. Higher numbers of patients and more investigation is required to establish whether this SNP is of any significance in epilepsy patients treated with VPA.
EXAMPLE 5 SAS Genetic Marker
[0150] 121 patients were analysed using the CASECONTROL procedure. rs2808526 from the GABBR2 gene (also known as GPR51) was the top SNP with a genotypic probability of 0.0001 and a genotypic chi-squared value of 18.25. Furthermore, the allelic probability was highly significant at 4.19E-05 and chi-square of 16.8, and the trend probability was 2.88E-05 with a chi-squared value of 17.5. Other SNPs that appeared highly significant across the three categories of tests were rs2229944 from the GABRB2 gene, rs507450 from the GRIA4 gene, and rs658624 from the SCN4B gene. The collection of top five SNPs is displayed in Table 9.
Table 9 SAS Genetic Marker Pharmacoresponsiveness analysis
Figure imgf000062_0001
[0151] The advantages of a function such as TPLOT are its capacity to bunch SNPs in linkage disequilibrium together. This is advantageous because the number of linkage blocks can be used as a better estimate in multiple hypothesis correction than individual SNPs. The output required to observe collinearity between SNPs in the large epilepsy dataset in this project is 8,166,861. Therefore, neither TPLOT nor Haploview has the capacity to analyze such large datasets. EXAMPLE 6 GeneRave
[0152] 115 patients were included in the GeneRave analysis. The HGmultc function was run to see which probsets of SNPs were chosen for responsiveness.
EXAMPLE 7 Supervised Learning
[0153] 115 patients were used in the building of the classifier, SNPs with multiple missing genotypes were removed, reducing the number of SNPs from 4,041 to 3,332. To remove selection bias of SNPs based on functionality, each SNP was allocated a unique number for the discovery/selection process. The top 15 SNPs for each of the cross- validated five independent cohorts and how they re-appear across the cohorts are shown in Table 10.
Table 10
Top 15 SNPs for each of the five independent validation cohorts. Arrows indicate a common SNP being picked up across the cohorts.
Figure imgf000063_0001
Figure imgf000063_0002
[0154] rs658624 and rs2808526 were found to be present in all five cohorts; rs4869682 was present in cohorts 1-4; rs3911833 was present in cohorts 1,3,4,5; rs678262 was present in cohorts 1,3,5; rs6001641 was present in cohorts 1,3,4; rs7153926 was present in cohorts 2,3,5; and rs2283170 was present in cohorts 3,4,5. Of these eight SNPs, rs6001641, rs3911833 and rs7153926 were removed for their low frequencies across genotypes, leaving five SNPs (rs658624 [SCN4B], rs2808526 [GABBR2 (GPR51)], rs2283170 [KCNQl], rs4869482 [SLCl A3] and rs678262 [SCN4B}). The k nearest number was chosen to be nine as nine neighbours produced the best detection rate in the cross-validated samples. Each validation cohort was then re-run through the classifier with these five SNPs exclusively. By restricting the number of SNPs to these five, each cohort produced an improved level of prediction compared to the original best five SNPs from 4,041. This indicates selection of these five SNPs has removed noise from other SNPs which were initially ranked as the top five for each cohort. Table 11 reports the Fisher t- test scores plus sensitivity, specificity and predictive value tests to measure classifier accuracy for each of the validation cohort predictors.
Table 11 Classifier predictive values
Figure imgf000064_0001
TP = Number of true positives: Responders correctly classified as responders. b FP = (type I error) Number of false positives: Non-responders incorrectly classified as responders. c TN= Number of true negatives: Non-responders correctly classified as non- responders. FN = (Type II error) Number of false negatives: Responders incorrectly classified as non-responders
* The actual p-value is 0.000000002.
[0155] The validation calculations of how predictive the classifier is, with an accuracy of 83.5% is provided below:
L = Ni+N2+N3+N4+N5
L = 25+24+22+21+23 = 115
Responders (R) = 83 Non-Responders (NR) = 32
Figure imgf000065_0001
1,2,3,4,5 L1 = (H5S) L2 = (17,7) L3 = (17,5) L4 = (15,6) L5 = (17,6)
[0156] For every v, a classification tree is constructed using training set (L- Lv). e.g. L] is classified based on training set (L2 +L3 +L4 +L5).
NK
Figure imgf000065_0002
i=l
Nv = Number of cases in Lv Rts (d(v)) = test sample estimate
(x,c) = x is attribute vector and c is class label i.e. (R/NR)
[0157] Test sample estimate for classification of cohort 1: 25
R's (d(/)) = 1/25 Σ itf'Xxd = C1), where Xj,Ci C Lv. i=l [0158] The resulting classification for each cohort is:
L1 = 22/25 L2 = 19/24 L3 = 18/22 L4 = 18/21 L5 = 19/23 [0159] Therefore the average of all the V test sample estimates:
5
Rcv (d) - 1/5 Σ R's (d(v)) i=l =l/5([22/25]+[19/24]+[18/22]+[18/21]+[19/23]) =0.834616 =83.5% accuracy
Drug specific pharmacoresponsiveness [0160] The 115 patients used in constructing the pharmacoresponsiveness classifier were sub-categorised into two drug types, carbamazepine and valproate. The number of patients remaining was insufficient to build a reliable classifier, with no cross-validating SNPs observed.
EXAMPLE 8 Replication
Novel classifier replication [0161] A subsequent prospective replication analysis of 45 patients from the newly treated epilepsy cohort (two of the original 49 had failed genotyping and two were not on AED therapy) showed the classifier predicted 28 of 29 (97%) responders correctly and nine of 16 (56%) non-responders (PPV = 80%, NPV = 90%, p = 0.0001).
SAS Genetic Marker replication
[0162] Within the replication cohort for the newly treated population, three SNPs used in the classifier appear to retain significance at an univariate level (Table 12). A one-tailed test was performed as the prior hypothesis in the application set was that differences would be in same direction as in the initial cohort. The highly ranked GABBR2 SNP from the initial analysis has a genotypic p-value of 0.03 and two SNPs located in the SCN4B gene have allelic and trend p-values of significance. SNP rs658624 has an allelic and trend p- value of 0.03. SNP rs678262 has an allelic p-value of 0.05 and a trend p-value of 0.03.
Table 12 Internal univariate replication analysis using SAS
Figure imgf000067_0001
EXAMPLE 9 Optimized training sets
[0163] A large number (111) of patients taking CBZ/VPA were used to validate the model with respect to generating an optimized training set. The optimized training set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3). Another set comprises rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs67826 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
[0164] The optimized training set provides 74% sensitivity and 52% specificity, with a p- value of 0.005 for predicting responsiveness.
EXAMPLE 10 Adverse Drug Reactions
[0165] The approach used to report the identification of a combination of SNPs that predict pharmacoresponsiveness in epilepsy patients to anti-epileptic drug treatment is also applicable to the identification of genetic predictors for adverse drug reactions (ADRs).
Adverse drug reactions occur in approximately 40% of patients commencing treatment with an anti-epileptic drug and can affect the neurological, immunological, metabolic or gastrointestinal system. The most common ADRs are: sedation, neurocognitive effects (especially poor concentration and memory), ataxia, weight gain, skin rash, bone mineral density loss and increase fracture risk.
[0166] The univariable/multivariate classification system is used aimed at validating a correlation between ADR and SNPs or other mutations. The mutations may be in one or more genes listed in Table 2.
[0167] In one embodiment, clinical samples are collected from a sufficient number of patients so that each ADR subgroup is represented by a n times wherein n is at least 1 ensuring that a statistically significant threshold is reached after application of the model.
[0168] Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications. The invention also includes all of the steps, features, compositions and compounds referred to or indicated in this specification, individually or collectively, and any and all combinations of any two or more of said steps or features. BIBLIOGRPHY
Akobundu et al, PharmacoEconomics 24:869-890, 2006
Begley and Beghi, Epilepsia 41:342-219, 2002
Binder et al, Clinical and diagnostic laboratory immunology 72:1353-1357, 2005
Bonferroni, Publicazioni del r Instituto Superiore di Science Economiche e Commerciali di irenze 8:3-62, 1936
Brennan, Am. J. Pharmacogenomics. J (4):395-302, 2001
Cariello Am. J. Human Genetics 42:126-134, 1988
Cotton et al, Proc. Natl. Acad. Sci. USA 57:4033-4037, 1988
Crimins et al, In Methods ofMicroarray Data Analysis (Springer røj: 191-205, 2005
Devlin and Roeder, Biometrics 55:997-1004, 1999
Edwards and Bartlett, Methods MoI. Biol. 22(5:287-294, 2003
Elghanian et al, Science 277:1078-1081, 1997
Fix and Hodges, Discriminatory analysis - non-parametric discrimination: Consistency properties (No. 4) Randolph Field, Taxas; USAF School of Aviation Medicine, 1951
Gianpiero and Weale, Lancet Neurology 6, 2007
Grompe et al, Proc. Natl. Acad. Sci. USA 55:5855-5892, 1989 Kim et al, BMC bioinformatics 5: 160, 2004
Kwan and Brodie, The New England Journal of Medicine 342:314-319, 2000
Kwok, Curr Issues MoI. Biol. 5(2):43-60, 2003
Lazarou et al, JAMA 279: 1200-1205, 1998
Mattson et al, The New England Journal of Medicine 575:145-151, 1985
Nakhaeizadeh and Taylor, Machine Learning and Statistics: The Interface (John Wiley & Sons, Inc):27-35, 1997
Novack et al, Proc. Natl. Acad. ScL USA 55:586-590, 1986
Orita, et al, Proc. Natl. Acad. ScL USA. 86:2166-2110, 1989
Sheffield et al, Proc. Natl. Acad. ScL USA 56:232-236, 1989
Shenk et al, Proc. Natl. Acad. ScL USA 72:989-993, 1975
Sidak, Journal of the American Statistical Association 62:626-633, 1967
Taylor et al, Biotechniques 30:661-669, 2001
Twyman and Primrose, Pharmacogenomics. 4(1):61-19, 2003
White et al, Genomics 72:301-306, 1992
Yeoh et al, Cancer CU 7:133-143, 2002

Claims

CLAIMS:
1. A method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, said method comprising screening for the presence of two or more mutations in genes selected from GABBR2, KCNQl, SCN4B and SLCl A3 which mutations correlate to potential responsiveness of the subject to the medicament.
2. The method of Claim 1 further comprising screening for a mutation in a gene selected from KCNC1/MY0D1, GRIA4 and GSTA4.
3. The method of Claim 1 or 2 wherein the mutations in the selected genes have a statistically significant linkage or association to a therapeutic outcome.
4. The method of Claim 3 wherein the response to the medicament is an adverse drug reaction.
5. The method of Claim 3 wherein the response to the medicament is a positive therapeutic reaction.
6. The method of Claim 1 or 2 wherein the mutation is single nucleotide polymorphism (SNP).
7. The method of Claim 6 wherein the SNPs are selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLC1A3).
8. The method of Claim 6 wherein the SNPs are selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rs!82623 (GSTA4).
9. The method of any one of Claims 1 to 8 wherein the subject is a human.
10. The method of Claim 9 wherein the neurological condition is epilepsy or related condition.
11. The method of Claim 10 wherein the medicament is an anti-epileptic drug (AED).
12. The method of Claim 11 wherein the AED is selected from the list comprising carbamazepine, valproate, phenytoin, lamotrigine and leviteracetam and analogs or homologs or classes thereof.
13. Use of two or more mutations in genes selected from GABBR2, KCNQl, SCN4B, SLCl A3 and SCN4B in the generation of a diagnostic assay to predict a therapeutic outcome of a drug proposed to be used for treating a neurological condition.
14. Use of Claim 13 further comprising the use of two or more mutations in genes selected from KCNCl /MYODl, GRIA4 and GSTA4.
15. Use of Claim 13 wherein the two or more mutations are SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), and rs4869682 (SLCl A3).
16. Use of Claim 14 wherein the two or more mutations are SNPs selected from the list comprising rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GSTA4).
17. A therapeutic protocol for the treatment of epilepsy or a related condition in a human subject, said method comprising determining whether said subject is a responder or non-responder to a particular AED by the method of any one of Claims 1 to 12 and then administering an AED to which the subject will respond.
18. A business method comprising inputting into a web-based site information concerning the presence of two or more mutations in an optimized training set of genes selected from mutations listed in Claim 15 or 16 wherein the web-based site provides an interactive response providing information on potential pharmacosensitivity or pharmacoresistance or an ADR to a drug used in neurological treatment.
19. A method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs in genes selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLC1A3) or rs3911833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4) and rsl82623 (GSTA4) which SNPs correlate to potential responsiveness of the subject to the medicament.
20. A method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, the method comprising screening for the presence of two or more SNPs in genes selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNC 1/MYODl), rs507450 (GRIA4) and rsl 82623 (GSTA4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLCl A3), rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl82623 (GST A4) which SNPs correlate to potential responsiveness of the subject to the medicament.
21. A method for a genotype-based prediction of responsiveness of a subject to a medicament in the treatment of a neurological condition the method comprising screening for the presence of two or more SNPs selected from the list comprising rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B) and rs4869682 (SLCl A3) or rs3911833 (KCNCl /MYODl), rs507450 (GRIA4) and rsl 82623 (GST A4) or rs2808526 (GABBR2 also known as GPR51), rs2283170 (KCNQl), rs658624 and rs678262 (both in SCN4B), rs4869682 (SLC1A3), rs391 1833 (KCNC1/MY0D1), rs507450 (GRIA4) and rsl82623 (GSTA4), which SNPS correlate to potential responsiveness of the subject to the medicament.
22. A method for determining the likelihood or otherwise of a subject responding or not responding to a medicament in the treatment of a neurological condition, said method comprising screening for the presence of two or more mutations in genes selected from the list set forth in Table 2, which mutations correlate to potential responsiveness of the subject to the medicament.
23. Use of two or more mutations in one or more genes in Table 2 in the generation of a diagnostic assay to predict a therapeutic outcome of a drug proposed to be used for treating a neurological condition.
PCT/AU2008/001556 2007-10-22 2008-10-22 A diagnostic assay WO2009052559A1 (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
AU2007905783 2007-10-22
AU2007905783A AU2007905783A0 (en) 2007-10-22 A Diagnostic Assay
AU2007906200A AU2007906200A0 (en) 2007-11-12 A diagnostic assay
AU2007906200 2007-11-12

Publications (1)

Publication Number Publication Date
WO2009052559A1 true WO2009052559A1 (en) 2009-04-30

Family

ID=40578959

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/AU2008/001556 WO2009052559A1 (en) 2007-10-22 2008-10-22 A diagnostic assay

Country Status (1)

Country Link
WO (1) WO2009052559A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010044683A1 (en) * 2008-10-15 2010-04-22 Nikola Kirilov Kasabov Data analysis and predictive systems and related methodologies
CN111462921A (en) * 2020-04-20 2020-07-28 山东大学 Epilepsy medication recommendation method and system
PL440701A1 (en) * 2022-03-21 2023-09-25 Warszawski Uniwersytet Medyczny Panel of markers for predicting the occurrence of epilepsy relapses in patients with tuberous sclerosis, in vitro method for predicting the occurrence of epilepsy recurrence in patients with tuberous sclerosis and use of a panel of markers in this method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1258495A1 (en) * 2001-05-09 2002-11-20 Millennium Pharmaceuticals, Inc. 98359, A sodium channel beta 4 subunit, and uses therefor
US20030162192A1 (en) * 2001-08-20 2003-08-28 Sotos John G. Polymorphisms associated with ion-channel disease
WO2004033722A2 (en) * 2002-09-23 2004-04-22 Sciona Limited Genetic profiling and healthcare management: adme (absorption, distribution, metabolism & elimination) & toxicology related genes and probes
US20050209181A1 (en) * 2003-11-05 2005-09-22 Huda Akil Compositions and methods for diagnosing and treating mental disorders
WO2008024677A1 (en) * 2006-08-21 2008-02-28 Novartis Ag Biomarkers for alzheimer's disease progression

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1258495A1 (en) * 2001-05-09 2002-11-20 Millennium Pharmaceuticals, Inc. 98359, A sodium channel beta 4 subunit, and uses therefor
US20030162192A1 (en) * 2001-08-20 2003-08-28 Sotos John G. Polymorphisms associated with ion-channel disease
WO2004033722A2 (en) * 2002-09-23 2004-04-22 Sciona Limited Genetic profiling and healthcare management: adme (absorption, distribution, metabolism & elimination) & toxicology related genes and probes
US20050209181A1 (en) * 2003-11-05 2005-09-22 Huda Akil Compositions and methods for diagnosing and treating mental disorders
WO2008024677A1 (en) * 2006-08-21 2008-02-28 Novartis Ag Biomarkers for alzheimer's disease progression

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
BALASUBRAMANIAN, S. ET AL.: "GABAB receptor association with the PDZ scaffold Muppl alters receptor stability and function.", JOURNAL OF BIOLOGICAL CHEMISTRY, vol. 282, no. 6, 2007, pages 4162 - 4171 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010044683A1 (en) * 2008-10-15 2010-04-22 Nikola Kirilov Kasabov Data analysis and predictive systems and related methodologies
US9002682B2 (en) 2008-10-15 2015-04-07 Nikola Kirilov Kasabov Data analysis and predictive systems and related methodologies
US9195949B2 (en) 2008-10-15 2015-11-24 Nikola Kirilov Kasabov Data analysis and predictive systems and related methodologies
CN111462921A (en) * 2020-04-20 2020-07-28 山东大学 Epilepsy medication recommendation method and system
PL440701A1 (en) * 2022-03-21 2023-09-25 Warszawski Uniwersytet Medyczny Panel of markers for predicting the occurrence of epilepsy relapses in patients with tuberous sclerosis, in vitro method for predicting the occurrence of epilepsy recurrence in patients with tuberous sclerosis and use of a panel of markers in this method

Similar Documents

Publication Publication Date Title
EP3730188B1 (en) Compositions and methods for use in combination for the treatment and diagnosis of autoimmune diseases
US20220033903A1 (en) Genetic markers associated with asd and other childhood developmental delay disorders
US11352672B2 (en) Methods for diagnosis, prognosis and monitoring of breast cancer and reagents therefor
US8431345B2 (en) Method for determination of progression risk of glaucoma
US8140270B2 (en) Methods and systems for medical sequencing analysis
WO2010042831A2 (en) Diagnosis, prognosis and treatment of glioblastoma multiforme
JP2009528063A (en) Markers for addiction
US20230304094A1 (en) Genomic alterations associated with schizophrenia and methods of use thereof for the diagnosis and treatment of the same
WO2010048497A1 (en) Genetic profile of the markers associated with alzheimer&#39;s disease
US8712696B2 (en) Methods of prognosing a rheumatoid arthritis remission phenotype
US20130296182A1 (en) Variability single nucleotide polymorphisms linking stochastic epigenetic variation and common disease
WO2011043333A1 (en) Susceptibility gene for hair shapes
US20070065821A1 (en) Methods for the prediction of suicidality during treatment
WO2009052559A1 (en) A diagnostic assay
WO2008010082A2 (en) Diagnostic method for fibromyalgia (fms) or chronic fatigue syndrome (cfs)
Smit et al. BEGAIN: a novel imprinted gene that generates paternally expressed transcripts in a tissue-and promoter-specific manner in sheep
KR20230005816A (en) Compositions and methods for evaluating the efficacy of neurotransmitter transporter inhibitors
US20140045717A1 (en) Single Nucleotide Polymorphism Biomarkers for Diagnosing Autism
US20120309641A1 (en) Diagnostic kits, genetic markers, and methods for scd or sca therapy selection
Amstadter et al. Selected summaries from the XVII World Congress of Psychiatric Genetics, San Diego, California, USA, 4–8 November 2009
Tang et al. Diagnosis of patients with mucopolysaccharidosis type II via RNA sequencing
WO2009036513A1 (en) Diagnostic and therapeutic protocols
Puri Molecular genetics of the 1q23. 3 schizophrenia susceptibility locus

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 08842478

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WPC Withdrawal of priority claims after completion of the technical preparations for international publication

Ref document number: 2007905783

Country of ref document: AU

Date of ref document: 20100407

Free format text: WITHDRAWN AFTER TECHNICAL PREPARATION FINISHED

Ref document number: 2007906200

Country of ref document: AU

Date of ref document: 20100407

Free format text: WITHDRAWN AFTER TECHNICAL PREPARATION FINISHED

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
122 Ep: pct application non-entry in european phase

Ref document number: 08842478

Country of ref document: EP

Kind code of ref document: A1