AU2005200652A1 - A method for selecting oligonucleotides having low cross hybridization - Google Patents

Description

I

AUSTRALIA

PATENTS ACT 1990 DIVISIONAL APPLICATION NAME OF APPLICANT: Combimatrix Corporation ADDRESS FOR SERVICE: DAVIES COLLISON CAVE Patent Attorneys 1 Nicholson Street Melbourne, 3000.

INVENTION TITLE: "A method for selecting oligonucleotides having low cross hybridization" The following statement is a full description of this invention, including the best method of performing it known to us: A METHOD FOR SELECTING OLIGONUCLEOTIDES HAVING LOW CROSS HYBRIDIZATION FIELD OF THE INVENTION The present invention is in the field of biological and chemical synthesis and processing. The present invention relates to methods for selecting oligonucleotides for low cross hybridization. The present invention may be applied in the field of, but is not limited to, the field of chemical or biological synthesis, diagnostics and therapeutics.

The present application claims priority to U.S. Provisional Patent Application Serial No. 60/116,956 filed January 25, 1999.

BACKGROUND OF THE INVENTION Advances are emerging continually in the field of biological and chemical processing and synthesis. Many novel and improved solid phase arrays or "gene chips" are being developed providing rapid methods for synthesizing chemical and biological materials. Examples of such technologies include those described by Pirrung et al., U.S. Patent No. 5,143,854, those described by Southern in WO 93/22480, those described by Heller in WO 95/12808 and those described by Montgomery in PCT/US97/11463. Hence, it is possible to synthesize, to manipulate and to examine ever increasing amounts of genetic materials. Moreover, it is possible to work simultaneously with, to analyze, and to test ever larger amounts of genetic materials.

Oligonucleotides can hybridize or bind to other oligonucleotides depending upon whether or not their sequences are more or less complementary. Sometimes, it is desirable to find a set of oligonucleotides that, as much as possible within a given set of constraints, do not hybridize or bind to each other. Optimization methods can be used to assist in selecting such a set of oligonucleotides.

There are many possibilities of how to make an oligonucleotide or oligonucleotides from a longer oligonucleotide. For example, one may cut the long oligonucleotide or select a segment or segments from which to form a smaller oligonucleotide. These smaller oligonucleotides formed by these and other methods known to those skilled in the art may be referred to as "substring sequences".

There is great flexibility on which substring of the oligonucleotide to select as a target sequence for binding.

Mitsuhashi et al., U.S. Patent No. 5,556,749 describe a computerized method for designing optimal DNA probes. The method is intended to produce probes designed for diagnosis and monitoring. However, the method described therein does not contemplate choosing more than one probe for simultaneous use with another probe whereby interaction and cross hybridization is minimized.

It is an object of the present invention to provide a method for choosing complementary substrings from among target oligonucleotides such that the substrings bind relatively well to their target oligo-ucleotides but do not substantially bind to other oligonucleotides in a sample. For instance, given a long oligonucleotide a string ofRNA, mRNA, DNA or cDNA) it is possible to select a piece or pieces from it for a later experiment, e.g. as a capture probe. In the case of an immobilized capture probe, it is desirable that the oligonucleotide substring sequence bind to the long oligonucleotide that it came from but not to other oligonucleotides that might be present in a sample.

It is often desirable to prepare an array of such immobilized capture probes so that each one binds or hybridizes to its intended target but does not bind or hybridize strongly to any of the other targets in the solution. An array of immobilized capture probes constructed according to the methods described allows using a smaller array of capture-probes for sequestering a desired set of targets because the array does not require as much redundancy for elucidating data points as other arrays where binding is not as clearly discernible.

SUMMARY OF THE INVENTION In a first aspect, the present invention features methods to provide a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets. The quality of hybridizing or binding relatively well to intended targets but not to unintended targets may be quantified using delta Tm (difference in strength of hybridization, such as difference in melting temperatures) according to the methods of the present invention. A probe having a relatively small delta Tm generally hybridizes to at least one unintended target substantially well. A probe having a relatively large delta Tm has substantially no unintended targets that it hybridizes to substantially well.

The methods according to the present invention feature: 1. Determining a set of targets. In some circumstances where it is not clear what the identity of all the targets in a particular solution might be, it is possible to determine a list of some of the targets that might be in the particular solution and to include that list in the set of targets. If there are some targets in the list that are not actually in the solution, it does not harm the quality of probes selected according to the present methods.

2. Selecting a particular target from the set of targets. This becomes the current target.

3. Choosing a sequence substring from the current target and providing its complementary sequence. This is a candidate probe. Choosing a sequence substring may be done by starting at a particular point in the current target and then incrementing the starting point by some amount each time a new substring is chosen, with wraparound of the starting point to the front portion S of the current target when an increment would otherwise run off the end of the target. A substring may be chosen at random or any arbitrary function may be applied in order to determine which substring to choose. If there are no more substrings in the current target that have not already been tried, it is recommended to: use the best candidate probe selected so far noting that this target might not be as selectively captured as desired, (ii) to do no more picking of probes for this target, and 1 (iii) to return to step 2, supra. Otherwise, using the candidate probe as determined and proceeding to lO step 4, infra.

O

O 4. Determining whether the candidate probe satisfies any criteria required or desired for jgf the set of probes. If it does not meet such criteria, returning to step 3 and choose a new candidate 0 o probe. If it does satisfy such criteria, proceeding to step Calculating the Tm for the candidate probe using the hybridization model.

Calculating substantially all possible cross Tm's of the candidate probe hybridizing to all unintended targets and finding the maximum cross Tm. Calculating delta Tm. Note that the set of all unintended targets will include previously picked probes if probes are also to be in solution as opposed to affixed to a support.

6. Proceeding to step 7 if delta Tm is acceptably large. Returning to step 3 and choosing a new candidate probe if delta Tm is not acceptable large.

7. At this point, a suitable probe has been chosen. The foregoing procedure may be repeated additional times to choose other probes for additional targets or, if desired, additional probes for a target for which one has already found one or more probes.

In a second aspect, the present invention features a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets.

In a third aspect, the present invention features a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets produced in accordance with the methods set forth, supra.

In a fourth aspect, the present invention features a programmed computer system for providing the sequences of a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets.

DETAILED DESCRIPTION OF THE INVENTION The present invention features methods to provide a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets. The quality of hybridizing or binding relatively well to intended targets but not to unintended targets may be quantified using delta Tm according to the methods of the present invention. A probe with a small delta Tm generally hybridizes to at least one unintended target relatively well. A probe having a large delta Tm has substantially no unintended targets that it hybridizes to relatively well.

The methods according to the present invention feature: 1. Determining a set of targets. In some circumstances where it is not clear what the identity of all the targets in a particular solution might be, it is possible to determine a list of some of the targets that might be in the particular solution and to include that list in the set of targets. If there are some targets in the list that are not actually in the solution, it does not harm the quality of probes selected according to the present methods.

2. Selecting a particular target from the set of targets. This becomes the current target for which one desires a probe.

3. Choosing a sequence substring from the current target and providing its complementary sequence. This is a candidate probe. Choosing a sequence substring may be done by starting at a particular point in the current target and then incrementing the starting point by some amount each time a new substring is chosen, with wraparound of the starting point to the front portion of the current target when an increment would otherwise run off the end of the target. A substring may be chosen at random or any arbitrary function may be applied in order to determine which substring to choose. If there are no more substrings in the current target that have not already been tried, it is recommended to: use the best candidate probe selected so far noting that this target might not be as selectively captured as desired, (ii) to do no more picking of probes for this target, and (iii) to return to step 2, supra. Otherwise, using the candidate probe as determined and proceeding to step 4, infra.

4. Determining whether the candidate probe satisfies any criteria required or desired for the set of probes. If it does not meet such criteria, returning to step 3 and choose a new candidate probe. If it does satisfy such criteria, proceeding to step Calculating the Tm for the candidate probe using the hybridization model.

Calculating substantially all possible cross Tm's of the candidate probe hybridizing to all unintended targets and finding the maximum cross Tm. Calculating delta Tm. Note that the set of all unintended targets will include previously picked probes if probes are also to be in solution as opposed to affixed to a support.

6. Proceeding to step 7 if delta Tm is acceptably large. Returning to step 3 and choosing a new candidate probe if delta Tm is not acceptable large.

7. At this point, a suitable probe has been chosen. The foregoing procedure may be repeated additional times to choose other probes for additional targets or, if desired, additional probes for a target for which one has already found one or more probes.

In particular embodiments where the probes and targets are both in solution the probes are not tethered to a support), it is preferable to include each previously accepted probe into the list of targets before calculating delta Tm for the current candidate probe. Those skilled in the art will readily understand that it is preferable to do so because the probes are able to hybridize to each other as well as to the targets and so should be included in any calculation of the strength of unintended S hybridizations. Those skilled in the art will readily understand that this feature means that the present invention provides probes having reduced cross hybridization with each other. Therefore, the present invention is particularly applicable to instances where multiple probes are used simultaneously in solution. This is often the case where primers are being used such as to amplify or copy sections of genetic material. In that case, "probe" and "primer" are meant to refer to the same oligonucleotide.

Adding targets as the methods according to the present invention progress may interfere with the ability to find a probe for a particular target not because of conflict with other targets but because of conflict (too strong a hybridization) to previously picked probes. If performing the methods outlined in the present invention reveals that there are a relatively large number of unacceptable probes S found, and if these probes are unsuitable because they bind too strongly to other probes, it is generally preferred to begin the method from the start using a different order of picking current targets. This generally results in a different order of picking probes and can result in a set of probes that are more mutually exclusive. For instance, if the first time, probes for targets are picked in the following order: target 1, target 5, target 2, target 4, target 3, and the best probes for target 2 and target 3 do not have an acceptable delta Tm because they bind relatively well to probes for target 1 and target 5, it is possible to repeat the methods according to the present invention again by picking probes for targets in for example, the following order: target 2, target 3, target 5, target 4, target 1. In effect, the sequence for choosing targets may be modified within the present methods. Alternatively, it is possible to start with the set of probes found so far, look at the probe that was found to be less than ideal, find what other probes it hybridizes too strongly with, redo those probes, and repeat this process until a compatible set of probes is found. By way of example, suppose that a set of probes is found, but probes 2 and 5 do not have an acceptable delta Tm because they hybridize too strongly with probes 1 and 3, respectively.

It is possible to eliminate probe 1 and proceed according to the methods of the present invention again for target 1 leaving the rest of the probes as is, finding a new acceptable probe for target I that perhaps does not conflict with probe 2. Then a skilled artisan may do the same for probe 3. This process (of redoing the previously found probes that later are found to conflict with other probes) maybe iterated until a good solution is found. If no suitable probes are found, one or more targets may be eliminated from the solution of interest.

In a second aspect, the present invention features a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets. The quality of hybridizing or binding relatively well to intended targets but not to unintended targets may be quantified using delta Tm. A probe having a small delta Tm generally hybridizes to at least one unintended target relatively well. A probe having a large delta Tm has substantially no unintended targets that it hybridizes to relatively well. Therefore, the set of probes according to the present invention may have, for example, a delta Tm of 50 C, 100 C or, for greater separation, 200 C. In general, the larger the delta Tm, the easier to dehybridize unintended targets while maintaining the intended targets hybridized by the probes in the subject set of probes.

In a third aspect, the present invention features a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantiallyto unintended targets produced in accordance with the methods set forth, supra.

In a fourth aspect, the present invention features a programmed computer system for providing the sequences of a set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets. Such a programmed computer system may comprise any one of a large number ofpossible software programs that may be designed by those skilled in the art without undue experimentation. Such a programmed computer system comprises a means for determining or designating one or more particular targets from a set of targets to probe (a current target), a means for determining or designating a sequence substring from the current target and determining its complementary sequence (a candidate probe). The means for choosing a sequence substring may function by starting at a particular point in the current target and then incrementing the starting point by some amount each time a new substring is chosen. A substring may be chosen at random or any arbitrary function may be applied in order to choose which substring to pick by the computer means, a means for determining whether the candidate probe satisfies any criteria required or desired for probes, and a means for calculating the Tm for the candidate probe using a hybridization model.

As used herein, the following terms are understood to mean the following: A "target" is an oligonucleotide in a sample.

A "probe" is an oligonucleotide intended to bind or hybridize to a target. Note that in cases where probes and targets are in solution, a particular oligonucleotide can be both a probe and a target.

A "set ofprobes" is intended to include two or more probes. Preferably, a "set of probes" includes or more probes. More preferably, a "set of probes" includes 100 or more probes. Even more preferably, a "set of probes" includes 1000 or more probes.

The "intended target" of a probe is the target that it is designed to best hybridize to.

Generally, this will be the target from which the probe is a complementary substring. For example, if GATTACAGATTACA is a particular oligonucleotide in solution (or target), one possible substring is S CAGAT. The complement of the substring CAGAT is ATCTG, and thus ATCTG can be a probe for O hybridizing to the intended target GATTACAGATTACA.

3 An "unintended target" is a target other than the intended target.

fl "Cross hybridization" is hybridization of a probe to a target other than its intended target or to 3 another probe.

"Tm" is most preferably the melting temperature of hybridization of a probe to its intended target. Melting temperature is defined in scientific literature and is used herein to describe a measure of how strongly a probe hybridizes to a target. More generally, Tm may be any useful measure of the strength of hybridization including, but not limited to, measures such as the best percentage match of the probe against a target, where A matches T and G matches C; the energy of binding of the probe against its target; the negative of the entropy of binding; some combination of the energy of binding and the entropy of binding; the enthalpy of binding; etc.

"Cross Tm" is the melting temperature of hybridization of a probe to an unintended target (or to another probe). For melting-temperature models that are location dependent, it is preferable to use the location where the melting temperature of hybridization is highest. Or, as in the description of Tm above, may more generally be some other measure of the strength of hybridization (such as percentage matching, energy of binding, negative entropy of binding, enthalpy of binding, combinations of these, etc.).

"Constraints" are the conditions or qualifications that must be substantially met when choosing probes. In most instances, "constraints" refer to a feature or property of the probe. For example, it might be desirable to select only those probes having a Tm between about 50 0 and 60 o C or that do not contain more than three G's in a row. There may be any number of heuristics or constraints imposed by preferences on the practitioner to use the probe set. Some exemplary constraints include that all probes are a particular length 20 bases long or 30 bases long) or that all probes have Tm's within a particular range within 5' C of each other or within 20 C of a mean Tm for probes having a particular length). In the case of using probes as primers, there can typically be constraints that the probe must bind to the target sequence within a certain area (in order to do the priming task correctly). Other constraints might be that the probe is not to have G's and C's within the last four base positions of its 3' end, and so on.

"delta Tm" is the difference, for a particular probe, between Tm and the maximum cross Tm.

A "hybridization model" is a mathematical model by which one calculates an estimated Tm or cross Tm based on the probe, the oligonucleotide to which it hybridizes and possibly the position of the hybridization. The hybridization model may also require the input of solution concentrations or additional factors. An important feature of a "hybridization model" is that it provides an estimate of Tm or cross Tm algorithmically. There are many hybridization models discussed in scientific literature and are believed applicable within the scope of the methods of the present invention.

Likewise, additional custom hybridization models may be created.

"Acceptable delta Tm" is the smallest delta Tm that is determined to be acceptable for a probe to be accepted according to the methods of the present invention. For example, an acceptable delta Tm might be 5* C, 100 C or, for greater separation, 200 C. Likewise, an acceptable delta Tm might be chosen as any number in between. In general, the larger the delta Tm, the easier to separate unacceptable from acceptable probes. Similarly, the larger the delta Tm, the easier to dehybridize unintended targets while maintaining the intended targets hybridized.

As used herein the term "bind relatively well to intended targets" is understood to describe a feature whereby a probe does not separate from but rather remains hybridized to a target sequence under normal operating conditions. A preferred example of such a feature is a perfectly complementary probe that does not separate from but rather remains hybridized to a target sequence at temperatures under 800 C As used herein the term "does not bind substantially to unintended targets" is understood to describe a feature whereby a probe does not hybridize to target sequences other than those to which it possesses a high degree of complementarity under normal operating conditions. A preferred example of a such a feature is a perfectly complementary probe that does not separate from but rather remains hybridized to a target sequence at temperatures below 80" C. However, the same probe does not bind to or easily separates from a target sequence to which it does not possess a high degree of complementarity at temperatures greater than some temperature significantly below 800 C, such as greater than 70" C, greater than 650 C, greater than 600 C, greater than 50 C, etc. The feature is intended to include minimal hybridization that may be reversed by agitation or heating above room temperature.

DESCRIPTION OF THE PREFERRED

EMBODIMENTS

The following are provided purely by way of example and are not intended to limit the scope of the present invention.

EXAMPLE 1 Selecting oligonucleotides for low cross hybridization The following is a sample list of targets for designing probes. The probes were to be built on a DNA chip such as those used in 'accordance with the method described by Montgomery in WO 9 8/0122 1, attached to a layer on the surface of a chip some other substrate so that the probes themselves are not floating around in solution. Thus, we do not account for probes in the target set during operation of the method.

>giI67l77381gbIAW305385.1IAW3o5385 xv93h12.xl NCICGAPBrn53 Homo sapiens cDNA clone IMAGE:28261 19 3Y, mRNA sequence

GCCAGTCACATGCTACCTGCATTAAAGACAGCMCAGGTAMGGGGACTACATTA

'TrACCAAAC CT7GGCMnGGGAGATATACAGGTCCGAGGAACTCGTGTCTACTGCAGACGATGCAAT

TACCCCACCT

TCCTCCATACAGAATGrTAGGATGTCACTCCGGGGGTGAT=CTCCTCAAGT

TGTAGCCAA

CA I II IGTCCGTAACTGATIrTCAGGGCAAACA-TTCTGACATCnTCCTCCAGCTCAGTCTG

CCATGCCTT

GGATCG'TCGCTTCACACAGTAGCATAA

GC

AGCTCAAAGT

GAAAGTGTTGCGTCTrGGTATCCGGAATCCTCAGCCCAGTAGCAGC=nAGTCATTC

ACG'ITCATC

>gi167175901gbIAW305237.1jAw305237 xr79hl l.xl NCICGAPLu26 Homo sapiens cDNA clone IMAGE:2766405 mRNA sequence

ACTACTATACGGCTGCGAGAAGACGACAGAAGGGTCATGTGUCTATATCACATA

TGGTITTGGAA

CCATCACCCCAAGGTA AA A AA AA AATAAAAGGTA'ErCCAGGTATGGGCATA

AATAAAGGTAA

'TAAAAACCGAAAA AA AAA AAAA AAAAA AAAA AAAA AA AAGTCGTATCGATGT >gil67141071gbIAW3O4418. I AW304418 xv6Ofl 2.xlI NCICGAPLu28 Homo sapiens cDNA clone IMAGE:2817551 3T, mnRNA sequence CTElCCTTATCACTCCCAGCAGATCMCT1CCTGTAGCflACCAC'CTAAA

TITTAATAT

GTGTTGAGCTCATrATTTAAGGAATCACATCnrGCTATCACATCCAGGCACCGGA

ACATAGTGTC

TATACTGACTGAACAGGCCAAGCTTCGTGAGTTAAAA

GGTAAGAACG

GTCCATCAT

ATCTTATCACTTGAGATGACAATGTrGAAJCTACAGGATGGAGGCATCTCAnTATfC

AGACCATTT

AAATCAATTATGACTTACAGTCTGTATACATAATCGGCCATATCA

AAAACTGAA

CCCAGTGGAAAATAmATATGTCCAAATAGGAGAGGAGTATAGCATGfl

GGTAAAT

>gil67137071gbIAW3O418.1jAw3o4ol8 xvl5hIl1.xI SoaresNFLTGBCSI Homo sapiens cDNA clone IAGE:28 13253 T, mRNA sequence TITITACAGGATAATAC1AATAGAGCACAUTATGTATCATTGGACCAGC

AAACAGACAG

TAGCAGGAAGGATGCTGCTCCCAAGGCTCTTCAGTCATCAGAGGACACACTCAAGCCCC

ACCTGAGTGT

TCTCCCCATTCCATCGGCCATCCCTGCTCAGGATGTGGTACCAGGGCCATCCCAACAGCC

TCATCTCAG

TAGACTCCAG=rGTCTAATTCTCCTrCAATGGTGCTCCTTGTCAC~CTCGTGGGCTGGC

GGATAG

>giI67135O71gbjAW3O3818.1IAW3o3818 xr23d05.xI NCICGAPUt4 Homo sapiens cDNA clone IMAGE:2760969 3T, mRNA sequence GAACTFGAATGTGCTEATATGCCACArCCCAGGAGAAAGAATAGTATCT

GAACAGGAA

TAATAATTCACAAATACTAACACTTATTGACAATAGACAAGTC=TAGGGTAGTGCA

CATGTACITrA AAAACTACCTCTACCAATCTCAACACThTATMpCAGTGAAACTGTAGCAGAT CCTAC1TA T=l~CAATGGTTAGTGTAAAAT[CTGTATGTAATAJAGTACATA IIm IGAGATGGAAGA

AGGACTGCA

TGTGAAATGTTGCCTAAGTGTAGGCTCCTGTCTACGCTATCAAGGCCAA

AAATCACTG

TITGGAGACAGAGMCGCTCTCTrGCCCAGGCTGGAGTACAJATGGCGCATCTCAG

CTCACCG

>gil67128981gbIAW3O3218. I AW303218 xr59g03.xI NCICGAP_0v26 Hiomo sapiens cDNA clone EIMAGE:2764468 3' similar to contains Mlu repetitive elemnent;, mRNA sequence

TATACGGCTGCGATAAGACGACAGAAGGGGTAGGACTGAGGCCTGAGTACACCTTAT

AYI=GGACAT

TACGTATrAAAAAAATfATCTAGCTGGGCATGGTGGCACACACCTATGGTCCCAGCTGC

TFFGGGAGGCT

GAAGTAGGAGGCTGGCTGAGCCCAGGAGmPAGTCCAGCCAGAGCACATAGTGAGA

ATTCATCTCAA

GA AAAAA AAA AGA AAA AAAA AA AA AGAAAAAAGTCGTATCGA >gil67l 18951gbJAW3Q2218.1IAw3o2218 xsQ3dO5.xl NOCICOAP_-Kidi I Homo sapiens cDNA clone JMIAGE:2768553 3' similar to TR:Q14934 Q14934 NF-AT3. mRNA sequence

CATATTACTGGTCATGAGCAGTMATGGGAGCAATCTGACCCCAGGTGCCAGCACAAJ

CAGCCAGCCC

ACACTCTAGACACGCCTTCACTCCAGTCCAJ'r'CTGGCACCTAGCCTCAGTCUTCACCCTCC

TCCCTCCTC

CACACACTCCTfCCCCCAGCCCTCCAAGGCAGCACCAGGCCTGAGGGCCACACCTCAGCT

GGGGGAGGGG

AGGGAAGACAGTGAGACAGACAGAAGCTGGGGAGAGGAGCCAGGGTGGCCCCAGG

CTTCTGTAGCCA

CCACTCCAGGAAGGAGGGAAAGGGGGCAGGGCTGAGGCTGGGGCTGGGGTTGCCAGGTG

ATGACAGTITCA

CGTGGTTCAGGCAGGAGGCTCTCTCCAGGAGGTGCAGGGAAGCCACTCAGGTCTCGGCC

AATGATCTCA

CTCACTGTAAAGGAGGGGCAGTTGAGAGACTGGGCTA

>gil67106951gbIAW301018.1 JAW30I11 xkl le~l.xl NCICGAPCo2O Homo sapiens cDNA clone IMAGE:2666424 mRNA sequence GCC I I I CAAGAATT~TII

CTAAMGTA

AACAGTAA

CAAAATAT[ACATAAAATAAAT[AACATGCAATrACTAACCATATGTATAA

CGTFTGGAATA

TATTAGCCTTCCCATGAGTTAATAAAAACTAATATrFGGT-TAGATTCATACCATCCT

TTCAAATAT

TrGGGTATGAAACT-FGGTAGCAATGCAATrrGTCTGATGTACAGAGCAGATrTCACCATGA

GAGATJ'ACAC

CAAAGAACAGATGTCCCTCCCAGAACATrATCTCACCCCAGACTCAGAAJACTGAGCAGC

CAAGCTTCCT

TCCCAGGAATCACCATGGAATGTCTGAACAATAACCAGGCCCTGGAGAflACTGCAGGGC

TGGCAGAGTT

TTAGGAATCAGCCAAACTC

>gil67104951gbIAW300818.11AW300818 xkO6eO9.xl NCI_-CGAP_-Col19 Homo sapiens cDNA clone IMAGE:2665960 3' similar to TR:088814 088814 HEME-BINDING PROTEIN. mRNA sequence AATATC= T AT CGAACAGTTAAGACCGT

AAAGATCAG

GCACAAATCACATTCCCCCT[rAATAACAAATACAATCCAATAA=1AGAUAATCA

GTIT=AGTG

ACCGTCTGGAACGCGGT

CGTTTGAA=TC

TCAATGAGA-A

GGATGCTGCATATCTGGCTGTATfATTFCCTACCGTGAGAAAGAGACTAGTATATGG AACATGC2T

MCGAATGATATATTAGGAGTGTATGATGAC

GGAACTTGGG

AACCGCCTCTCCCAGAGAAAAAGAATCATAT

C'ITGGTTCAG

TGGCCCTTTCAACAACTATCCGCGAGCTAGG

GGGTCATAAC

CCTCGATGTTCNCGAGGGTTCCTAGGGAGA

>gij670545 81gbIAW298 822.1 JAW298822 UI-H-BWO-ajq-h-09-O-UI.s 1 NCJ CGAP Sub6 Homo sapiens cDNA clone IMAGE:2732800 mRNA sequence CGGCCGCGCCGGTTEE1TCAAGTFIGGGTATGTJT&JLTCTGTTATGTACTACTGTTCTG TTTGflTAT

CTCACAGTGC

TGTGTGCCCCGGTCACCTAGCAAGCTGCCGAACCAAAGAATHTGCACCCCGCTGCGGGC

CCACTTGGTT

GGGGCCCTGCCCTGGCAGGGTCATCCTGTGCTCGGAGGCCATCTCGGGCATAGGTCCACC

CCGCCCCACC

CCTCCAGAACACGGCTCACGCTTACCTCAACCATCCTGTF1'GCGGCGTCTGTCTGAACCAC

GCGGGGGCC

TTGAGGGACGCTTTGTCTGTCGTGATGGGGCAAGGGCACAAGTCCTGAATGTTGTGTGTA

TCGAGAGGCC

AAAGGCTGGTGGCAA

We used the following constraints. Probes must be 20 bases long. Probes must have a Tm within 10 C of the expected Tm for a 20-mer according to the hybridization model used (in this case 68.250 C).

We used the following hyridization model: Tm 81.5 0.41 Pgc 675 N Pmm, where Pgc is the percent GC content of the probe (number of G's number of C's) N 100, N is the length of the probe in bases, and Pmm is the percent mismatches (number of mismatches) N 100.

We chose an acceptable delta Tm of 200 C.

The algorithm worked as follows. We began with target 1. We picked a 20-mer out of it at a randomly selected location and found its complement as a candidate probe. We checked that the candidate probe satisfied the constraints. If not, we chose another 2 0-mer from a random location. If it did, we then calculated the Tm's for this probe hybridizing to all other targets at all other locations and used that data to find the maximum cross Tm and thus delta Tm. If delta Tm was greater than or equal to 200 C, we kept this probe and obtained a probe for the next target (target If not, we chose another 20-mer from a random location. We repeated this process until we found one acceptable probe for each target.

The following is the list of probes found by this process. In the following, there is header information given for each probe indicating from which target it comes what its intended target is), where in that target the probe comes from at what offset into the intended target), the Tm of the probe, the maximum cross Tm of the probe, what unintended target provides the maximum cross Tm, and where in that unintended target the maximum cross Tm happens (at what offset).

Note that the Tm values given all match exactly. The experimentally determined Tm's will not necessarily match exactly the Tm's given are estimated Tm's derived from the hybridization model, which in this case results in the methods described being able to find probes that all match exactly in estimated Tm.

probe 1 from target 1 at offset 359; Tm 68.3; max. cross Tm 33.3 from target 9 at offset 253

ATTCCGGATACCAAGACGCA

probe 2 from target 2 at offset 2; Tm 68.3; max. cross Tm 48.3 from target 6 at offset -3

CTTCTCGCAGCCGTATAGTA

probe 3 from target 3 at offset 145; Tm 68.3; max. cross Tm 28.3 from target 5 at offset 272

AAGCTTGGCCTGTTCAGTCA

probe 4 from target 4 at offset 239; Tm 68.3; max. cross Tm 23.3 from target 1 at offset 176

AAGTGACGAAGGAGCACCAT

probe 5 from target 5 at offset 424; Tm 68.3; max. cross Tm 28.3 from target 1 at offset

GAGCGAAACTCTGTCTCCAA

probe 6 from target 6 at offset 36; Tm 68.3; max. cross Tm 28.3 from target 1 at offset 225

AAAGGTGTACTCAGGCCTCA

probe 7 from target 7 at offset 333; Tm 68.3; max. cross Tm 28.3 from target 1 at offset 314

ACGTGAACTGTCATCACCTG

probe 8 from target 8 at offset 352; Tm 68.3; max. cross Tm 28.3 from target 5 at offset 24

CATTCCATGGTGATTCCTGG

probe 9 from target 9 at offset 210; Tm 68.3; max. cross Tm 33.3 from target 4 at offset 79

AGCCAAGATATGCAGCATCC

probe 10 from target 10 at offset 350; Tm 68.3; max. cross Tm 28.3 from target 1 at offset 175

ACAGACAAAGCGTCCCTCAA

EXAMPLE 2 In this example, the list of targets was the same as in Example 1. Likewise, all of the parameters and the model used for calculating Tm were the same as in Example 1. The only difference was that we used a variation of the method.

We began with target 1. We picked a 20-mer out of it at a randomly selected location and found its complement as a candidate probe. We determined that the candidate probe satisfied the constraints. If not, we picked another 20-mer from the "next" location, supra. If it did, we calculated the Tm's for this probe hybridizing to all other targets at all other locations and used that data to find the maximum cross Tm and thus delta Tm. If delta Tm was greater than or equal to 200 C, we kept this probe and moved on to getting a probe for the next target (target If delta Tm was not greater than or equal to 200 C, we went back and picked another 20-mer from the "next" location (see below).

We repeated this process until we found one acceptable probe for each target.

By "next location," we applied the following process. We selected a new candidate probe starting at a location one base to the right (in the 3' direction) of the previous pick. If such a location resulted in not having enough bases to make a candidate probe (such as when the next location is too close to the end of the target so that there are not enough bases left to make a probe of the desired length), we started at location 1 of the target. Thus, the process of scanning a target for an acceptable probe was started at a randomly selected point and then progressed incrementally along the target with wrap-around to the front of the target when the end was reached.

This process provides an exhaustive search of a target for an acceptable probe. It will find an acceptable probe if one exists. Thus, it is a good candidate search method for situations where the targets might be very similar except for small differences (perhaps mutations) at particular sites in the oligonucleotide.

This process resulted in the following set of probes being found.

probe 1 from target 1 at offset 62; Tm 68.3; max. cross Tm 33.3 from target 3 at offset 372

CCCAAAGCCAAGGTTTGGTA

probe 2 from target 2 at offset 0; Tm 68.3; max. cross Tm 38.3 from target 6 at offset

TCTCGCAGCCGTATAGTAGT

probe 3 from target 3 at offset 115; Tm 68.3; max. cross Tm 33.3 from target 6 at offset 173

TATGTTCCGGTGCCTTGGAT

probe 4 from target 4 at offset 234; Tm 68.3; max. cross Tm 33.3 from target 1 at offset 399

ACGAAGGAGCACCATTGAAG

probe 5 from target 5 at offset 292; Tm 68.3; max. cross Tm 33.3 from target 7 at offset 111

GGAGCCTTACAACTTAGGCA

probe 6 from target 6 at offset 154; Tm 68.3; max. cross Tm 33.3 from target 3 at offset 73

TTAAACTCCTGGGCTCAAGC

probe 7 from target 7 at offset 265; Tm 68.3; max. cross Tm 28.3 from target 4 at offset 93

AGTGGTGGCTACAGAAGCTT

probe 8 from target 8 at offset 379; Tm 68.3; max. cross Tm 28.3 from target 1 at offset 59

ATCTCCAGGGCCTGGTTATT

probe 9 from target 9 at offset 438; Tm 68.3; max. cross Tm 38.3 from target 4 at offset 194

CGCAATGAGATCTGGCTGTT

probe 10 from target 10 at offset 350; Tm 68.3; max. cross Tm 28.3 from target 1 at offset 175

ACAGACAAAGCGTCCCTCAA

Although the invention has been described with reference to the presently preferred embodiments, it should be understood that various modifications can be made without departing from the spirit of the invention. Accordingly, the invention is limited only by the following claims.

Claims (10)

1. A method to provide a set of probes that hybridize relatively well to their intended targets but do not substantially hybridize to unintended targets comprising the steps of: Determining a set of targets; Determining a particular current target from the set of targets to probe; Choosing a sequence substring from the current target and providing its complementary sequence, which becomes the candidate probe; Determining that a candidate probe satisfies any criteria desired or required for probes; Calculating the Tm for the candidate probe using a hybridization model; Calculating substantially all possible cross Tm's of the candidate probe hybridizing to all unintended targets and finding the maximum cross Tm; Calculating delta Tm; Determining whether the delta Tm is acceptably large. Repeating steps forward until the desired probes are found.

2. The method of claim 1 wherein choosing a sequence substring is performed by starting at a particular point in the current target and then incrementing the starting point each time a new substring is chosen by some amount.

3. The method of claim 1 wherein the substring is chosen at random.

4. The method of claim 1 wherein the delta Tm is at least about 200 C. The method of claim 1 wherein the delta Tm is at least about 100 C.

6. The method of claim 1 wherein the delta Tm is at least about 5' C.

7. A set of probes that hybridize or bind relatively well to their intended targets but do not bind substantially to unintended targets.

8. The set of probes of claim 7 wherein the delta Tm of the set is at least about 20' C. Q:\OPER\fx\005Lcb\ 12567990 divfil 045.doc. 1/0205 -16-

9. The set of probes of claim 7 wherein the delta Tm of the set is at least about The set of probes of claim 7 wherein the delta Tm of the set is at least about 5 0 C.

11. A set or probes that hybridizes or binds relatively well to intended targets but do not bind or substantially hybridize to unintended targets produced in accordance with the method of claim 1.

12. A programmed computer system for providing the sequences of a set of probes that hybridizes or binds relatively well to intended targets but that does not substantially hybridize or bind to unintended target comprising a software program having a means for determining or designating one or more particular targets from a set of targets to probe (a current target), a means for determining or designating a sequence substring from the current target and determining its complementary sequence (a candidate probe), a means for determining whether the candidate probe satisfies any criteria required or desired for probes, and a means for calculating the Tm for the candidate probe using a hybridization model. DATED this 14 th day of February, 2005 COMBIMATRIX CORPORATION by its Patent Attorneys DAVIES COLLISON CAVE