CA2381685A1 - Method of non-targeted complex sample analysis - Google Patents

Abstract

A method for non-targeted complex sample analysis which involves the following steps.
A first step involves providing a database (16) containing identifying data of known molecules. A second step involves introducing a complex sample containing multiple unidentified molecules into a Fourier Transform Ion Cyclotron Mass Spectrometer (12) to obtain data regarding the molecules in the complex sample. A third step involves comparing the collected data regarding the molecules in the complex sample with the identifying data of known molecules in order to arrive at an identification through comparison of the molecules in the sample.

Description

~ ~
~ ~ ; .
METHOD. OF NON-TARGETED COMPLEX SAMPLE ANALYSIS
FIELD OF THE INVENTION:
The present invention relates to a method of non-targeted complex sample analysis, with particular application to biology, and genomics in particular.
BACKGROUND OF THE INVENTION
Functional genomics is an emerging field in biotechnology that focuses on the characterization of gene function. All organisms contain only one genotype.
However, the expression of this genotype under varying developmental and environmental conditions results in an almost infinite number of possible phenotypes. It is the correlation of gene expression to phenotype that defines functional genomics. To properly study a gene we need to not only know its identity (a.e. sequence) but to be able to observe and characterize its expression patterns in response to developmental and environmental changes, in isolation as well as in relation to the other genes in the genome. To properly study the effects resulting from the expression of a gene we need to be able to characterize the phenotype resulting from this activity in an objective and quantifiable manner.. This is what the non-targeted metabolic profiling technology invention described herein enables the functional genomics community to do.
The gene sequences of entire species are now known: Gene-chip technology has made it possible to monitor and quantify the changes in expression of each and every genie within the genome to developmental and environmental changes, simultaneously.
Gene-chip technology is, in essence, non-targeted gene expression analysis even though it is, in actuality, a targeted analysis that just so happens to contain all of the possible targets. This is a powerful comprehensive capability, but it was made possible by the fact that the genome is a finite and unitary entity. The analogous phenotypic capability would be to have every metabolite and protein of an organism known and on a chip. This is not possible due to the fact that not only are there multiple phenotypes, but a virtually infinite number of metabolites and proteins are possible. To be complementary to the current state of genomic analysis, phenotypic analysis must be non-targeted in "actuality".
The non-9 v.f targeted metabolic profiling technology described herein is the only platform that satisf es the requirements of non-targeted phenotypic.analysis. Furthermore, this technology is not restricted to any one species, but is equally effective in all plant and animal species.
Deciphering the complex molecular makeup of an individual phenotype is a formidable task. To be able to accurately and reproducibly generate this phenotypic information in such a way that the virtually infinite number of possible.phenotypes can be compared to one another and correlated to gene expression is the crux of the dilemma that faces functional genomics. On the molecular level, the phenotype of a given biological system can be divided into the proteome and the metabolome. Since gene expression results in protein synthesis, the proteome is the first and most direct link to gene expression. However, due to the complex interactions of metabolic pathways, it is difficult to predict the effects that changes in the expression levels of a given protein will have on the overall cellular processes that. it may be involved in. The metabolome, on the other hand, is the summation of all metabolic (proteomic) activities occurring in an organism at ~ any given point in time. The metabolome is therefore a direct measure of the overall or end effect of gene expression on the cellular processes of any given biological system at any given time. For this reason, the metabolome should prove to be the more powerful of the two phenotypes in actually understanding the effects of gene function and manipulation. The non-targeted metabolic profiling technology described herein is the only comprehensive metabolic profiling technology available.
Isolation, identification, and quantitation are the three fundamental requirements of all analytical methods. The primary challenge for a non-targeted metabolome analysis is to meet these requirements for all of the metabolites in the metabolome, simultaneously. The second and perhaps more difficult challenge is to be able to meet these requirements with sufficient throughput and long-term stability such that it can be used side by side with gene-chip technology. Such technology will drastically reduce the time that is required for the function of a particular gene to be elucidated. In addition, databases of such analyses enable very large numbers of phenotypes and genotypes to be objectively and quantitatively compared. There is no such product or technology available to functional genomics scientists at this time. The non-targeted metabolic profiling technology described herein has been extensively tested in multiple species. In all cases, the ~i technology has verified the metabolic variations known to exist between various genotypes and developmental stages of different species.
Key Technology Concept. The non-targeted metabolic profiling technology.
described herein can separate, quantify and identify all of the components in a, complex biological sample quickly and simultaneously. This is achieved without any a priori selection ofthe metabolites of interest and is therefore unbiased. These data are exported to a database that allows the researcher to directly compare one sample to another (l.c. mutant vs. wild-type, flowering vs. stem elongation, drought stress vs. normal growing conditions;
etc.) or to organize the entire database by metabolite concentration (l.c. which genotype has the greatest or least expression of a given metabolite). This technology is equally applicable to the study of human disease. To make use of this information, the researcher just types in the empirical formulas) or the accurate masses) of the metabolites) he or she is interested in and the software will organize the data accordingly.
The ability to conduct an analysis of the composition of substances in biological samples is critical to many aspects of health care, ~ environmental monitoring as well as the product development process. Typically the amount of a specific substance in a compleX
mixture is determined by various means. Far example, in order to measure analytes in a complex mixture, the analyte(s) of interest must be separated from all of the other molecules in the mixture and then independently measured and identified.
In order to separate the analytes in a complex mixture from one another, unique chemical and/or physical characteristics of each analyte are used by the researcher to:
resolve the analytes from one another. These unique characteristics are also used to identify the analytes. In ;all previously published reports of complex mixture analysis, the methodologies require known analytical standards of each potential analyte before the presence and/or identity of a component in the.unknown sample can be determined: The analytical standards) and the unknown samples) are. processed in an identical manner through the method and the resulting characteristics of these standards recorded (for example: chromatographic retention time). Using this information, a sample containing unknown components can be analyzed and if a component in the unknown sample displays j the same characteristic as one of the known analytical standard (s), the component is postulated to be the same entity as the analytical standard. This is targeted analysis technology. Targeted analysis technology is one-way: The researcher can go from known standard to methodology characteristics but not from methodology characteristics to known standard. The researcher can only confirm or refute the presence and/or amount of one of the previously analyzed standards: The researcher cannot go from the method characteristics of an unknown analyte to its chemical identity. The major drawback of this type of analysis is that any molecule that was not identified . prior to analysis is not measured. As a result, much potentially useful information is lost to the researcher. To be IO truly non-targeted the method must allow the researcher to equally evaluate all of the components of the mixture, whether they are known or unknown. This is only possible if the defining physical and/or chemical characteristics of the analyte are not related to the method of analysis but axe inherent in the composition of the analyte itself (i.e. its atomic composition and therefore its accurate mass):
Key benefits of non-tar~~eted metabolic profiling technology 1. Multidisciplinary. Virtually only one set of analyses would need to be performed on a given sample and the data resulting from this analysis would be available to all scientists regardless of the area of research they are focusing on. .
2. Comprehensive. The non-targeted approach assesses ALL metabolite changes and will thus lead to a faster,and more accurate determination of gene function/disfunction.
3. Unknown Metabolite Discovery. The non-targeted approach has the, potential of identifying key metabolic regulators that are currently unknown, and which would not be monitored in a targeted analysis scenario.
4. High Throughput. The system is can be fully automated and analysis time is short allowing 100's of samples to be analyzed per instrument per day.

5. Quantitative. The system is reproducible and has an effective dynamic range > 104.
Relative changes in metabolite expression over entire populations can be studied.
Business Impact of Technolo~y. The ability to generate searchable databases of the metabolic profiles.of a given organism will represent a revolution in how the effects of genetic manipulation on a species can be studied. Currently our knowledge of the actual genetic code is much greater that our knowledge of the functions of the genes making up this code. After the mapping of the genome, the next greatest challenge will be determining the function and purpose of these gene products and how manipulation of these genes and their expression can be achieved to serve any number of purposes. The 5 time, energy, and cost of investigating the effects of genetic manipulation are great. A
database that can be searched for multiple purposes and which contains direct measures of the metabolic profiles of specific genotypes has the potential to dramatically decrease the amount of time required to determine the function of particular gene products.
Such a database will reduce the risk of investing a large amount of time and resources. researching genes which rnay have effects on protein expression, but due to down-stream feedback mechanisms, no net effect on metabolism at the whole cell or organism level.
In an article published in CURRENT OPINION IN PLANT BIOLOGY in 1999 entitled "Metabolic Profiling: a Rosetta Stone for genomics?", Trethewey, Krotzky and Willmitzer indicated that exponential developments in computing have opened up the "possibility" of conducting non-targeted experimental science. While recognizing that it would not be possible to. work with infinite degrees of freedom, the opinion was advanced that the power of post-experimental data processing vi~ould make possible this non-targeted approach.
The non-targeted approach described in that article dealt only with the post acquisition analysis of metabolite data; not the non-targeted collection of metabolite data.
Thus the feasibility of non-targeted analysis of complex mixtures is neither obvious nor simple. The three major problems surrounding the non-targeted analysis of complex mixtures are: the ability to separate and identify all of the components in the mixture; the ability to organize the large amounts of data generated from the analysis into a format that can be used for research; and the ability to acquire this data in an automated fashion and in a reasonable amount of time.

y j SUMMARY OF THE INVENTION
What is required is a method of non-targeted complex sample analysis.
According to the present invention there is provided a method for non-targeted complex sample analysis that involves the following steps. A first step involves providing a database containing identifying data of known molecules (this database contains the elemental compositions of all molecules previously identified in nature, organized by species, metabolic processes, subcellular location, etc.). A second step involves introducing a complex sample containing.multiple unidentified molecules inta a Fourier Transform Ion Cyclotron Mass Spectrometer to obtain data regarding the molecules in the complex sample. A third step involves comparing the collected data regarding the molecules in the complex sample with the identifying data of known molecules in order to arrive at an identification through comparison of the molecules in the sample.
Molecules that are not represented in the database (i.e. .unknowns) are automatically identified by determining their empirical formula. Thus; the method allows rapid identification of new molecules within the complex mixture related to specific molecules already identified, as well as identification of those molecules within the complex mixture that bear no relationship to those class or category of molecules already defined. ~s a.
result the analysis of complex mixtures is greatly simplified.
The invention, as described, uses the high resolving power of Fourier Transform Ion Cyclotron Mass Spectrometry (FTMS) to separate all of the components within the mixture tFiat have different empirical formulas. This has been shown for petroleum distillates, but not for aqueous biological samples ionized in a "soft"
ionization mode, where adduct ions can be problematic. The accurate mass capability of FTMS
that enables the determination of empirical formula has been widely established.
Furthermore FTMS is capable of performing high resolution/accurate mass 2D MS/MS which provides structural information that can be used to confirm the identities of components that have identical empirical formulas and allows the organization of metabolites based upon common structural components. This capability has been shown by isolated research groups but is not available on a commercial instrument. By integrating these capabilities with an automated sample injection system and an automated data integration and database system, all of the components within a complex mixture can be analyzed rapidly and simultaneously. The data is then exported into a database that can be searched and organized by sample, or analyte. It is to be noted that unlike the approach advocated by Trethewey, Krotzky and Willmitzer, the present method is not dependant upon the advances in post experimental data processing. The non-targeted metabolic profiling technology described herein generates a dataset that is simple and compact.
Computing technology capable. of organizing and interpreting the described databases is readily available. No new advances are required: Furthermore, the' technology does not have the finite limits inherent in the approach of Trethewey; Krotzky and Willmitzer.
BRIEF DESCRIPTION OF THE DRAWINGS
These and other features of the invention will become more apparent from the following description in which reference is made to the appended drawings and figures, the drawings and figures are .for the purpose of illustration only and are not intended to in any way limit the scope of the invention to the particular embodiment or embodiments shown, wherein:
FIGURE 1 is a side elevation view depicting non-targeted analysis of complex samples in accordance with the teachings of the present invention.
FIGURE 2 is an illustration of raw data (mass spectrum) collected from the FTMS
showing how the metabolites in the complex mixture are separated from one another. Mass range displayed 100-350 amu.
FIGURE 3 is an illustration of raw data (mass spectrum) collected from the FTMS
showing how the metabolites in the complex mixture are separated from one another. 10 amu mass range displayed.
30' FIGURE 4 is an llustration of raw data (mass spectrum) collected from the FTMS showing how the metabolites in the complex mixture are separated from one another. 1 amu mass range displayed.

FIGURE 5 is an illustration of raw data (mass spectrum) collected from the FTMS
showing how the metabolites in the complex mixture axe separated from one another.
Mass range displayed 100-350 amuØ1 amu window.
FIGURE 6 is an illustration of strawberry pigment pathway (comparison of different developmental stages of an organism).
FIGURE 7 is a~ illustration of the extracted mass spectra of Phenylalanine from strawberry extracts from different developmental stages.
FIGURE 8 is an illustration of the extracted mass spectra of Cinnamate from strawberry extracts from different developmental stages.
FIGURE 9 is an illustration of the extracted mass spectra of 4-Coumarate from strawberry extracts from different developmental stages.
FIGURE 10 is an illustration of~the extracted mass spectra of Naringenin from strawberry extracts from different developmental stages.
FIGURE 11 is an illustration of the extracted mass spectra of Pelargonidin from strawberry extracts from different developmental stages.
FIGURE 12 is an illustration of the extracted mass spectra of Pelargonidin-3-glucoside from strawberry extracts from different developmental stages.
FIGURE 13 is an illustration of glucosinolate mutants in Arabidopsis thaliana (comparison of genetic mutants to wild-type and identification of unknown metabolites).
Relative changes in 3-Methylthiobutyl Glucosinolate illustrated.

FIGURE 14 is an illustration of glucosinolate mutants in Arabidopsis thaliana (comparison of genetic mutants to wild-type and identification of unknown metabolites).
Relative changes in 3-Methylsulphinyl~ropyl Glucosinolate illustrated.
FIGURE 15 is an illustration of glucosinolate mutants in Arabidopsis thaliana (comparison of genetic mutants to wild-type and identification of unknown metabolites).
Relative changes in 3-Methylsulphinylheptyl Glucosinolate illustrated.
FIGURE 16 is an illustration of Tobacco Flower Analysis (Location of metabolite expected to be responsible for red color in tobacco).
FIGURE 17 is an illustration of Tobacco Flower Analysis (Location of unknown metabolite potentially involved in tobacco color).
FIGURE 18 is an illustration of Observed Metabolic Changes in Strawberry Development.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
The preferred method of non-targeted complex sample analysis embodiment will now be described with reference to FIGURE 1 The purpose of this invention is to provide a means of analyzing large numbers of complex samples, for example biological extracts, and be able to analyze the information in a non-targeted fashion after-the analysis is complete to determine the differences between samples.
In the invention complex samples are directly injected into the FTMS 12 though the use of an autosampler 14 with or without the additional use of a chromatographic column. The components of the mixture are ionized by one of many potential "soft'' ionization sources (electrospray, APCI, FAB, SIMS, MALDI, etc.) and then transferred into the ion cyclotron resonance (ICR) cell with or without additional mass-selective pre-separation (quadrupole, hexapole; etc.). The ions are then separated and measured in the ICR cell with or without simultaneous MS/MS occurring. The data collected (mass spectrum) is integrated (the mass, relative intensity, absolute intensity of each ion is determined) and processed, with or without calibration with known molecules of known concentrations. These data, with or without isotope elimination and empirical formula calculation, are then transferred to a database 16 that organizes and stores the data for future comparisons and .functional 5 analyses. Once stored in the database, individual samples can be compared with one another and those molecules that show different concentrations between the selected samples can be displayed: The entire database can be searched for specific molecules.
The samples in the database can be listed from highest to lowest concentration or vice-versa. The molecules detected in the analysis can be compared with a database of known 10 molecules and the molecules automatically identified. For molecules that do not match known molecules, the most likely empirical formulas can be displayed.
This approach provides numerous advantages to the researcher. There is a dramatic increase in the amount of information obtained from each sample (> 1 Ox compared to the most comprehensive targeted analysis procedure reported). Information is collected on both known.and unknown components of a mixture. There is increased e~ciency of data collection (data collection is approximately 10x faster than reported targeted analysis techniques). It provides a basis for unbiased comparison of unknown samples.
Effects of gene modification on total cell metabolism can be determined instead of effects on only a small subset of metabolic processes (i.e. the relationship between different metabolic processes can be studied). By analyzing all metabolites the actual step within a metabolic process that is disrupted can be determined. Gene modifications that have an effect on protein expression but no net effect on cell metabolism can be identified. All of these analyses are completed simultaneously in one fast analysis; whereas multiple tirile-consuming analyses would have to be performed to get identical data at a tremendously higher cost Many examples exist for the use of FTMS for the analysis of complex mixtures, but none have introduced the concept of non-targeted analysis followed by database formation. The described method recognizes and utilizes some heretofore, unused capabilities in FTMS.
FTMS has the theoretical resolving power to separate all of the metabolites of different empirical formula in a . coriiplex biological sample. FTMS has the theoretical accurate n mass capabilities to assign empirical foimulas to all of the metabolites in the complex biological sample. FTMS has the capability to perform 2 dimensional MS/MS on all of the metabolites in a complex biological sample. It is not necessary to know a priori what metabolites are present in a complex biological sample if the analytes could thus be S separated and then be identified based upon their empirical formula and MS/MS fragment data and or by comparing them to a database of known analytes. Complex samples can be compared with one another to determine what analytes had different intensities. between the samples. A database could be organized by analyte or by common MS/MS
fragments.
This approach significantly decreases the time and resources needed to elucidate gene function as a result of genetic manipulation, environmental changes, or developmental changes in an organism. One of the many applications of the described method invention include gene function determination in functional genomics research.
Numerous targeted LC-MS methods as well as other screening methods have been I5 developed to analyze specific molecules or groups of molecules in complex samples. The major reason that this invention is novel and not obvious is because it employs a fundamentally different strategy for analytical analysis and is only possible with highly specialized instrumentation and methodology. Although the many independent theoretical research capabilities of FTMS have been known for at least 10 years, FTM$ has only been used in a targeted way and for specialized research purposes. In the past 10 years no group has described the application of FTMS employed within the scope of the present invention.
The present invention involves the combining of several theoretical FTMS
capabilities into a comprehensive, non-targeted metabolic profiling procedure that has commercial utility in the analysis and interpretation of complex mixtures.
The method of the present invention comprises the following steps:
Generation of Known Metabolite Database. The identity (common name and empirical formula) and relevant biological, information (species, 'metabolic processes involved in, cellular and subcellular location, etc) of all known biological metabolites are inputted into a commercial database program (i.e. Microsoft EXCEL, Table I:). The accurate monoisotopic mass of these metabolites is automatically determined along with their [M+H]+ and [M-H]- accurate mass (M+H and M-H refer to the mass of the metabolite when a proton (H+) is either added to the metabolite to create a positively charged ion or removed from the metabolite to create a negatively charged metabolite). The data collected from the FTMS analysis of the complex sample can then be compared to this database to immediately identify many of the components in the complex sample.
Preparation of samples for analysis: The metabolites are extracted from their biological source using any number of extractionlclean-up procedures that are typically used in quantitative analytical chemistry. Procedures are normally tailored to the source of the.
sample (i.e. leaf tissue, root tissue; blood, urine, brain; etc). For example, a 0.1 g plant leaf sample may be extracted by placing it, 1.0 ml of 50/50 MeOH/0.1 % formic acid, and 3 small glass beads; in a test tube and then vortexing for one minute to homogenize the sample. The test tube is then centrifuged for 5 minutes. 100u1 of the supernatant is then transferred from the test tube. to a 96 well plate. The 96 well plate is placed upon the autosampler. 20u1 of the supernatant is injected into the FTMS.
Typical operating conditions Solvents. 50/50 MeOH/0.1 % ammonium hydroxide as the mobile. phase and for dilution for all negative ionization analyses and 50/50 MeOH/0.1 % formic acid for all positive ion analyses.
Instrmrientation. Brisker Daltonics APEX III Fourier Transform Mass Spectrometer (FTMS) equipped with a 7.0 Tesla actively shielded super conducting magnet with electrospray (ESI) and atmospheric chemical ionization (APCI) sources. ESI;
APCI, and ion transfer conditions were optimized for sensitivity and resolution using a standard mix of serine, tetra-alanine, reserpine, , HP Mix, and adrenocorticotrophic hormone fragment 4-10. Instrument conditions were optimized for ion intensity and broadband accumulation over the mass range of 100-1000 amu. One megaword data files were acquired and a sinm data transformation was performed prior to Fourier transform and magnitude calculations.
Calibration. Ali samples were internally calibrated for mass accuracy over the approximate mass range of 100-1000 amu using a mixture of the above~mentioned standards Sample Analysis Samples are introduced to the FTMS via an autosampler, or in some cases with a syringe pump. When the sample solution reaches the source of the FTMS (the source is where the FTMS ionizes the molecules in the sample solution), then molecules are ionized according to the principles of the particular ionization source used. The source can either be external to the mass analyzer or internal, depending on the type of ionization (for example in ESI
and APCI ions are generated external to the mass analyzer and then transferred to the mass analyzer, whereas in electron impact ionization the molecules are ionized internal to the mass analyzer). The ions once generated and ~ transferred (if necessary) to the mass analyzer are then separated and detected in the mass analyzer based upon their mass to charge ratio.
Analyte Detection All of the analytes within the complex mixture are analyzed simultaneously (see Figures 2-5). Structurally specific information (accurate mass with or without accurate MS/MS
fragment masses) is obtained for all of the analytes without prior knowledge of the analyte's identity; and then this data is forrilatted in a way that is amicable to a comprehensive database.
Complex Sample Database Formation -The typical process of database formation involves the following steps:
1. The output of the FTMS (calibrated mass spectrum) is filtered to remove aII
13C isotopes and peaks that have mass defects that do not correspond to singly charged biological metabolites;
2. Each of the peaks. in this filtered peak list is then analyzed using the mass analysis program that is part of the instrument manufacturer's software package according to the elemental constraints provided by , the researcher. This program returns all of the possible elemental compositions that are possible at a given mass within a certain selected error range.

3. Only the data (file name, sample ID, mass, relative intensity, absolute intensity, empirical formula(s)) from those peaks in the filtered peak list that satisfied the above constraints are exported to a final processed data file (Table II). Each sample: analysis results in such a final processed data file.
S 4. Multiple databases can then be formed from the combining arid comparing of the data files. Three such databases are:
a) Direct comparison of two samples to create a database of differences (Table VI);
b) Combination of multiple files to create a database capable of tracking changes through a series of samples (Table III);
c) Direct comparison of a whole series of samples to one control sample and then the combination of all the samples in the series into one database to allow comparisons within the series vs a common control (Figure 8).
The utility of the invention is illustrated in the following examples:
I The ability to compare different developmental stages of an organism (Figures 6-12.
Table IV
In this example, v~%e looked at the strawberry pigment pathway in strawberries. Figure 6 shows the full metabolic pathway. Figures 7-12 show the various metabolites in the pathway that we observed. It is to be noted that we were able to Look at molecules of vastly different chemical compositions (amino acid, acid, flavenoid, glucoside). Here we were able fo see the changes within a single genotype (red strawberry) as a function of developmental stage (green - white - turning - red) and compare it to a different genotype (white mutant). Only the non-targeted metabolic profiling technology described herein has this broad of a spectrum. Furthermore, as indicated in Table IV, these changes in the metabolome are directly correlated with changes in gene expression.
II. The ability to compare different genotypes (Figures 13-.15, Table V).
In this example three different Arabidopsis thaliana mutants (TUI, TU3, TUS) that are knowwto have changes in the content and concentration of glucosinolates were compared IS
to a wild-type (WT). In this instance the non-targeted metabolic profiling technology described herein was 'able to. confirm previous results as well as identify glucosinolate changes that had never before been observed.
III. The ability to detect and identify unknown metabolites involved in ke~pathways (Fig_ures 16 and 17, Table IX).
In this example the flowers of a control (red): tobacco was compared to a white mutant. It was expected that the glucoside (Figure 16) was the metabolite responsible for color.
However, when analyzed by the non-targeted metabolic profiling method, the expected metabolite was not observed. An unknown metabolite (Figure 17) was detected and identified. (Table IX) to be the metabolite responsible for tobacco flower color.
IV The ability to compare the effects of different environmental conditions on an org.,anism ~ able VI) In this example the exuate from a carrot root grown under normal growing conditions (sufficient.phosphate) was compared to the exuate from a carrot. root grown under abnormal growing conditions (insufficient phosphate). Using non-targeted metabolic profiling we were able to identify key plant hormones that are excreted to~
promote symbiotic fungal growth under conditions of low phosphate.
V. The ability to group and classify metabolites based upon accurate MS/MS
data (Table VII and Table VIII) In this example accurate MS/MS fragmentation data was collected on the metabolites that were observed to be increased in the low phosphate conditions described above.
Classes of molecules that have a similar substructure can be grouped together (in this case all metabolites with the C l OH9N602 fragment). This capability greatly enhances the ability to search and characterize different complex mixtures VI: The ability to comprehensively monitor the metabolites of an organism (Table X, Fi ure 18 In our study of the developmental stages of strav~berry, we characterized the number of metabolites that we were observed as well as the number of metabolites that were observed to have changed in concentration between the different developmental stages.
It is the comprehensive nature of this method that allows one to monitor and evaluate virtually ail ongoing metabolic processes independently or in relation to one another. No other technology has this capability.
Table I Example of Known Metabolite Database Monoisotopic Masses Common Natne Metabolic AbbreviationC H N S M M+H M-H
Process O
P

glyoxylate 2 2 3 74.000475.0076 72.9931 Glycine GIy,G 2 5 1 75.032076.0392 74.029F

pyruvicacid PA 3 4 3 88:016089.0233 87.008f L-Alanine AIa,A 3 7 I 89.047790.0549 88.0404 Lactic Acid 3 6 3 90.031791.0389 89.024!

Cytosine . 3 5 3 99,0432160.0505 98.036( Acetoacetic acid 4 G 3 102.0317103.0389 101.024:

gammaaminobutytate GAGA 4 9 I 103.0633104.0705 102.0561 L-serine 3 7 1 105:0426106.0498 104.035 Histamine 5 9 3 111:0796t 12.0869 110.072 .

Ucacil 4 4 2 112.0273t 13.0345 111.020( 3-cyanoalanine 4 6 2 114.0429113.0501 113.035:

L-Proline - Pro.P S 9 I 115.0633116.0705 114.0561 L-Valine VaI,V 5 111 117.0790118.0862 116.071:

succinate 4 6 4 118.0266119.0338 1t7.019~

L-Homosedne 4 9 1 119:0582120.0655 1t8.081( L-Tbreonine Thr,T 4 9 1 119.0582120.0655 118.051(.-.

phosphoenolpyruvicacid PEP 3 6 3 1?1.0054122.0127 119.998:

L-cysteine . Cys,C 3 7 I I 121.0197122.0270 120.012:

Nicotinic Acid 6 S I 123.0320124.0392 122.0241 7ltytnine 5 6 2 126.0429127.0501 125.035:

L-Isoleucinc Ile,l 6 13I 131.0946132.1018 130.087 ' 2 L-Leucine Leu,L 6 13I 131.0946132.1018 130.087 oxaloaceticacid OAA 4 4 5 132,0059133.0131 130.998!

L-aspargine Asn,N 4 8 2 ! 32.0535133.060? 131.046:

L-Omithine 5 .1?o 132.0899133.0971 131.082( ?

L-Aspattate ' Asp,D 4 7 I 113.0375134.0447 132.030:

Ureidoglycine 3 7 3 133.0487134.0559 132.041:

L-malic acid 4 6 5 134.0215135.0287 133.014;

Ureidoglycolate 3 G 2 134.0327135.0400 133.025:

' L-Homocysteine 4 9 I I 135.0314136.0426 134.028:
?

Adenine (Vitamin , S 5 5 131.0545[36.0617 134.047:
84) Adenine 5 5 5 1_s5.0145136.0617 134.047:

3-MethyleneoxindoleAuxins 9 7 1 145.05?8146.U600 144.045.

Indolealdehyde Auxins 9 7 I 145.0528146.0600 144.045:

IndoleninetipoxideAuxins 9 7 1 145.0528146.0600 144.045' alphaKetogtutarate 5 G 5 146.02747,0287 145.014:
l5 LGlutamine GIn,Q 5 f02 146.0691147.0753 145.0615 L-Lysine Lys,L 6 142 145,1055147.! 127 145.098:

L-Glinamate ~ GIu,E 5 9 I 147.0531148.0604 146.0455 L-Methionine Met,M 5 III 1 149.0510150.0583 148.0431 D-ribose 5 105 150.0528151.0600 149.045( Guanine 5 5'S 111,0494152.0566 150.042:

lndole-3-acetotiuileAuxins IAN l07 2 155.0609136.0681 154,053:.

J

Comments: Any molecule of known chemical composition can be added to the database at . any time. The database is comprised of accurate monoisotopic masses. All molecules that have a unique empirical formula will, have a unique accurate mass. This mass is a constant and is independent of the methodologies discussed herein making it possible to analyze all of the components in a complex sample in a non-targeted fashion.
Figure 2 shows two raw mass spectrums. The top one is from the extract of a green stage strawberry and, the lower one is from the extract of a red ' stage strawberry.
Over 500 unique chemical entities were observed over the mass range displayed above (100-350 amu; which is only a subset of the entire mass range analyzed (100-5000)).
Figures 3, 4, and 5 show smaller and smaller mass ranges to illustrate the separation of the metabolites.
Figure 5 shows the resolution of the mass spectrum above 165,000. This extremely high resolution is necessary in order to separate all of the metabolites and thus be able to compare the two samples and determine the changes, if any.

Table II Illustration of processed data (file ID, mass, intensity, empirical formula, relative error) FdeID iv~ss fm C N O Err C H N 0 P S Err H P
S

99.044061 205E+06 5 0 2 0 0.05 6. 0 99.044082 9.33Ef06 5 0 2 0 0.26 1020549 256E+06 4 1 2 0 0.25 102054956 3.080-+96 4 1 2 0 0.01 102054962 1.36Ef06 4 1 2 0 0.07 104:0705 1.93E+06 4 1 2 0 0.10 104.079624 1.75E+06 4 1 2 0 0.18 104.106977 2.73E+06 5 1 1 0 0.13 104.106979 2.73E~06 5 1 1 0 O.t1 104.106981 1.84C-~065 1 1 0 0.09 104:107 3.88E+06 5 1 1 0 0.09 106.0498 1.21 E+08 3 1 3 0 0.01 106.04987 :1.36E~ 3 1 3 0 0.00 106.04987 1.630-+OB 3 1 3 0 0.00 106.04987 1.OBE+OB 3 1 3 0 0.00 106.04987 1:53E-rOB 3 1 3 0 0.00 106.04987 2:5 3 1 3 0 0.00 106.04987 2.45E+a8 3 1 3 0 0.00 106:04587 2:62E+08 3 1 3 0 0.00 106.04987 2.480-f08 3 1 3 0 0.00 106.04987 233E+OB 3 1 3 0 0.00 107.07Q2<i7 a.34E06 4 0 3 0 0.31 107.070322 1.28Ei06 4 0 3 0 0.48 108.080743 279E06 7 1 0 0 0.30 109:D28414 1.65E+06 6 0 2 0 0.07 111.044016 1.41 E+066 0 2 0 0.36 114.091316 274~E~6 6 1 1 0 0.21 114.081319 3.02E-r066 1 1 0 0.19 114.091336 1.76E+06 -6 1 1 0 :0.04 114.091337 3:87E~6 6 1'1 0 0.03 914.091342 270E+06 6 1 .10 0.01 114.091346 3.26E-r066 1 1 0 0.05 114.091358 3.18E-(066 1 1 0 0.15 114.091375 2740-+06 6 1 1 0 0.30 114.091377 2.53E+06 6 1 1 0 0.32 114:091404 2210-+06 6 1 1 0 0.56 115.038%8 3.43E~6 5 0 3 0 0.11 ' 115.038978 2:03E+06 5 0 3 0 0.07 115.008984 1:84+06 5 0 3 0 0.12 115.0999 1.57E~+fl6 5 0 3 0 0.25 115.039032 1;86E~(J65 0 3 0 0.53 115.03905 1.67E+06 5 0 3 0 0.69 116.034226 1.76E~06 4 1 3 0 0.06 116.034233 243E+06 4 1 3 0 0.12 116.03425 2070-+06 4 1 3 0 0.26 116.070538 2.600-+065 1 2 0 0.58 116.07060'( 1.46E~6 5 1 2 0 0.03 116.070643 1.46E-r065 1 2 0 0:33 118:086184 1,56E~16 5 1 2 0 0.60 118:086217 4:1~+Q6 5 1 2 0 0.32 .

118:066231 1.520-+065. 1 2.0 0.20 118.086r134 1.23E+065 1 2 0 0.18 118.086246 274E+06 5 1 2 0 0.08 118.096249 2.53E-r065 1 2 0 0:05 Comments: The h~ass spectrum is processed such that the 130 isotopes are first eliminated (this is only passi6le in FTMS analysis due to the high resolution and mass accuracy).

Then the remaining peaks are automatically analyzed using the mass analysis program that is included with the instrument using specific constraints chosen by the researcher (in the above example only those peaks that have the appropriate combination of carbon (C), hydrogen (H); oxygen (O), nitrogen (N), sulfur (S), or phosphorus (P) are returned). The final dataset now only contains monoisotopic, singly charged metabolites that have an accuracy of measurement of less than 1 ppm (err).
Table III Illustration of the database generated from the processed data;
Flnpiric~l (been Tutli~g i~d. St~e Fcnr>~a Sage Sage Write Stage C NO P S Int tv~ss U~t NESS 'fSCST5IY16fvdss RS11GS1~1~6ii~'i5 H tut~ss Vs6IGS i~ . Ink 21 D100 0 1.30006 ref 100433.113012921292433.1128 229232292'31774 20 rtF 1.30006 1.68E+07 . 2.9$E+08 25 6190 0 1.30E+06 723.19654008723.19528615215723.1853 10846271126 34 ~ 5.21EU7 1.12E+D8 1.41008 24 0130 0 1.30E~06 ~t 1005191132 243 243519.1133 930890083829 22 rd 1.30C-DDB . 3.16E+06 1.2'fE+OB

22 6. 0 0 1.30Et06 rf 100rd 1.30006100 100397.2714 486'148624862 32 1 rif 1.30E+06 6.32007 46 t11 0 1 1.30E06 790.2821-2.620072015790.28195.71E074382218790.2822 35 ref 4.54E+07 19 113 0 0 1.306-+06 448.152715448.1'.13754138448.15A2 309211482 17 td 3.53E+07 4.88E07 4.~E-~07 11 49 0 1 1.30006 381:07101292381.07101685130381.0709 2115164126 i6 of 1.68E~07 219E07 275E+07 9 85 0 3 1.30E-+06 of 100of 1.30E+D6100 100415:0638 206920692069 18 rB 1.30C-f96 2.~E~07 30 194 0 0 1.30E-+06 of 100758.57 25152515758.5698 1877187775 67 rd 1.30E+06 3:2'7E+07 2.44007 47 73 0 0 1.30006 782.56972823782.5694245487 78257 16315B 66 71 ~ 3.67E+07 3:19E+07 212E+07 22 145 0 2 1.30E+06 645.28251746695.28232065119645.2825 163193 78 40 of 2.27EE97 27100T 212C-+07 23 85 0 1 1.30Ef06 525.1667319525:16631185371525.1664 11~ 36699 24 rfi 4.15E~06 1.54E~07 1.52E+07 9 81 0 3 1.30E+06 rtf 100349.06831.42E~+06109 109349.06851.50Et07115411541056 16 of 1.30E+D6 20 4110 1 1.30C-+D6 533.1550442533.15511185268533.1550 106224090 28 of 5:75E-ro6 1.54E+07 1.38007 22 31 0 3 1.30006 448.15461031448.15451331129448.1546 101599 76 29 ~ rf 1.34007' 1.707 1.~0'l 33 69 0 0 1.300-+06 679.40311169679.40251215104679.4028 100886 83 54 . of 1.520Q7 1.58Ef07 1.310-+07 14 3130 0 1.300-X06 448:1774900448.17741177131448.1T14 965 10984 29 rd 1.170-+07 1.53E-~97 128E+07 D110 0 1.300-t06 ~f 100ref 1.30E06100 1003TI:1078 ~
r~ 1.30E+06 1:24Ei07 21 02 0 1 1.30E06 ~ 1.30006100of 1.30Ei06100 100329.06341.17E+Q7900 12 ~

40 80 D 3- 1.30006 r~F 100rif 1.300-+061D0 100723.2143 869 869869 34 of 1.30E+96 1.13E+07 27 25 0 2 1.30006 54Z:3240931547.3239936 101547.3240 ; 88 87 50 ~ 1:21E+U7 1.22E+U7 1.06E+07 815 21 2210 2: 1.30E-~6 ~ 1.300D6100- r~ ' 100725.1951 8013808808 44 r~ . 1.30E+06100 1.05Ef07.

0'170 1 385707.22161060275707.2218 792 20675 42 707.222203 5.3407 3.990-X07 5.04E+06 707.2220 1.94E+07 12 4110 1 1.30006 of 1.30E06100of 1.3QE+O6100 100433.1235 76;i763763 24 r~ 9.920D6 Comments: In Table III, the data was sorted according to the relative expression of metabolites in the ied stage vs the green stage of strawberry. The data can be organized by any field. What is observed is that the metabolite ClOH20010 has a concentration that is at least 22923% of that observed in the green stage (this metabolite is not observed in the 1 ~ green stage so the value is a % of the background noise). This metabolite can be identified by its empirical formula as pelargonidin-3-glucoside, the primary pigment observed in strawberries that give them their red dolor. This process is automated.

r Table IV Comparison of Metabolite and Gene Expression Data in Strawberry Color Formation (Red Stage vs. Green Stage) v Relative Relative l~Ietabolic Pathway Metabolite Gene Expression Expression 4-Coumarate-COA to Nargingenin 4.3 3.3 Chalcone Naringenin Chalcone to Naringenin4.3 4.3 Leucopelargonidin to Pelargonidin20* 6.7 Pelargonidin to Pelargonidin-3-Glucoside42* 8.3 *Reflects greater dynamic range of metabolic expression analysis Comments: Figures 7 through 12 and Table IV show the power of non-targeted metabolic profiling in studying changes that occur during development: Non-Targeted metabolic profiling allows the researcher to monitor entire metabolic pathways simultaneously.
There is no other methodology that allows for the simultaneous analysis o~
such a diverse 10 range of analytes. All of the analyte.s illustrated above were extracted from the non-targeted data collected using the methodology and concepts presented in this application.

i and identification of unknown metabolites). Relative changes in 3-Methylsulphinylheptyl Glucosinolate illustrated.
Table V Comparison of Glucosinolates in different Arabidopsis thaliana mutants Arabidopsis Glucosinolate ts lV~utan Glucos inolates R= WT : TU1 TU3 TU5 TU7 3-Methylthiobutyl 1:00 <0.06(nf)2.69 0.14 0.36 3-Methylthiopentyl 1.00 <0:56(nf)2.12 <0.56(nf)0.71 3-Methylthioheptyl 1.00 1.00 <0.21 0.32 e0.21 (nf) (nf) 3-Methylthiooctyl 1.00 2.93 <0.09(nf)0.92Ø15 3-Methylsulphinylp~opyl 1.00 27.62 1.37 21.560.37 3-Methylsulphinylbutyf 1.00 0.10 2.50 0:63 0.53 3-Methylsuiphinylpentyl 1.00 1.56 3.11 0.79 1.11 3-Methylsulphinylheptyl 1.00 1.38 <0.37(nf)0.64 ep.37(nf) 3-Methylsulphinyloctyl 1.00 6.16 e0:11 4.25 0.37 (nf) 3-Indolylmethyl 1.00 4.44 0.90 1.85 0.71 -Methoxy-3-indolylmethyl 1.00 1.41 0.67 0.59 0.46 C3H70S 1.00 >6.88 of of of (nf) C5H1108S 1.00 2.68 0.73 0.85 0.60 C7H1OOS3 1.00 >5.73 of >3.01of (nf) C8H120S3 1,00 <0.37(nf)1.95 <0.37(nf)0.45 C43H26N03S 1.00 2.55 1.05 1.18 0.44 C21 H2303 1.00 2.74 1.21 0.47 0:52 19 Glucosinotate Molecules Observed (17 reported) Comments: In Table V, the applicability of the technology for comparing genetic mutants to their wild-type counterparts is illustrated. The non-targeted metabolic profiles of four mutants(TUI, TU3, TUS; and TU7) were compared to their wild-type counterpart.
Here we show that not only can we identify and monitor the glucosinolates that had been previously analyzed using targeted analysis, but were able to identify previously unidentified glucosinolates. As is the case in all of our analyses, all of the other metabolites are also available for evaluation.

T able VI Ill~xstration of daiabase ger>erated by directi;y Co_~~~aring r~ro sc_~sples caxv°ot root equate in the presence arvd abs'vnce of piiosphate~ Slarnm.ary of ~:etaboi:ites that ~rere Observed to be Increased 1r°1 the -~ Fraction -P!+p I 1 F Minus P Pius P ~''~__ Pro xd Empirical Formula _ Observed ;-heoretical Erro;
Rat't.o iCorr,) I Mo~ ( Mass Abs Int. 1 Corr.lrtt. ~~ Mass -, Abs Int. ~ ~ C H
N I O~P S/C~ 1 Na;Ki a ) As _ Mass m _ 9972.550 _ j ESI+ _ 240783 1 2.358+09j 9.178+09 _____ _1.908+OG 70 9 5 ' 2 _! __~_.y I ~_-1 _ +H ~ 245.07895_ 0.73 1053.350 ~ ESI+ _ 467.7672 I 2.9?E+09 1.058+09 ~ 1.C~OE:-D6 _~ ~22 ~ 6 ~6 j '~, I ~ t ~~_ +H 467,16735891 -0.45 981.550 ~. .- ESI+ 777.0546 ~ 7.958+p9 9.528+08 1.008+p6 101. 1 3 ! I ~,-- +H
777.0546206 I -0. 57 658.650_':, 851+ ~ 223.9965 1.328+09 6.59~*08 ________,'~ODE+O6 i2 1~: i_, , ' : I 1 223.0564854 -O.7G
186.090 _ ESI+ 269.9524 3.728+08 1.868+08 I 9.008+OG _-~ 12', 14 ~: ~_ !, !I t ? __ +K _ 269.0523672 0.05' 73 375 i_ 851+ __ 6592412 l 9_478+08 T.34E+07 ~ ~~ 1.O~OEw06 -131 ' 3S' 6 90j I -i ~ +H 659.2409778 0.48 _ 52.845 _!_8S9+ 328.7390 1.0&8+08 5288+p7 ~ 1.998+06 75'22 5 7 ' I s j _! _ I
+H ! 328.1390785 -0.24 17.308 __ j ESi+ __ _699.2509 I 9.468+C7 4.738+07 _j _ 1.908+p6 831 35 6 ~ 8 I, I j . '' ~-1 __~_ +H ~' 619.2590885 -0.39 3S 421 _I _ ESI+ 559.3239 7.958+071 3.548+07 ' 1.008+O6 ~~.! 28 43 ~ 6 6 1 -i +H 559.3238595 0.93 __ _-3-~-. .1 _L.__.. ' __ ~~ . _ 34.279 ~ '. - ESI+ 539.2693 6.868+07 3.43~+07 ~~ ~OOE)U 1C6 27 ~ 35 6 6 , , .
j -1 +H 539.2612593 0.00 35.780 j ESI+ 307.0489 ; 6.368+07 1 3.988+07 _ 1.008+06 ~_ t2 19 j 3, ;_. 1 3 ~~.~-~ t _ +H 397.049083 -0.60 28.736 _ ESI+ _ _523._2299 '' S.63E+D7~ 2.8 0 _ i'~1,OOE+OG _ 26131 6 _j 6 ''.' y ~ 9 +H _ 523.20.'99592 -O.U9 25.570 _ ESI _ 569.?988 j 5.198+9712.558+07 ___,-- ____ 1,008+p6 __ 26 29 _G_ 9 ' 1_ -~, ~i-1 -_- +H 569.799053 ~0.44 _ x4.248 _ ~ ESI- 27S.12;~.42E+0712.428+C7 __ _ ~ 1.008+O6 _ 751 991 ~~ ~~ ;~~
! I 7 -H 279,9237973 -0.60 _ 22.393 ~5I+ 635.3554 4.488+07 2.248+07 ~ I 1.008+O6 34.47,6 1 5 ' I j i;-1 _ *H 635.3559597 0.36 _.._ _.__ _._.....__._.._.~.. _ . 1.
.-_27.312 ~. . ~~__ESI+'_-_ 543.3288 I 4.268-r07 2.938*07 ~, ~ .... 1 9.00_+DG
-._ - ~8~43' g -~..,~.. __ __. - øH __. 5,3.3289449 -0.2?
20.003-',__~~_.4PCIf 377.7594~._1.2.OCE+0712.CUc~07 --..1.__.~~-__._~~t.COE+06 - ~~ _20i25~__.. % : 1 -.. __._-9 ___ -*li-- 377.1594796 ~0.18 19.937 ~Sla 291.0714 : _3.998+07 -~. °98+07 -~._~_- I 1.00~+06 ~ 11 151 ' g~' ~ -~-j-_ ~ ~ _7 +H 291.0710585 9.03 __ 75374 : pPCI- 279. 1.538+071 9.638+07 ~ - "-- ~1.UOE+D6 _ '_5199~~-~ 5~_ _i , I 7 -H 279.9237973 0.26 __ ".3.32'? _- GSI+ -_- 467.2663 I 2.668;-071 Y."sac-e07 _'__~__-_ - - j1 JOE+06 _ j24:'35' 6 r5_' ~___;__~. ~ I-7 -__ +H 487.2663447 -O.G7 ~_ 73.273 1 SI ~ 335.2227 ; 6.63c=07~-6.638+07 _~- 335.2227 - S.DGE+OG '2D~I31 ; 4 ~; ~ 1 -H , 335.2227831 ~-0.40 I
1 __53.097 _ n APCI- - 335.2230 ' 1.608+08' 7.608+08 335.2231 7.228+07 ~ 20 31 4 I ~ ; 1 -H 335,2227831 0.66 F
-+ F+
_ 12.908 _~_ 851+ 242.9700 2.598;-47 5.908 07 ____1.0D_ O6 15 20 '!0 9 ~ ~ -~~
I -2 _ +2H 242.0791876 -0.85 3 X9.693 ~~ ESI+ _ 473.2507 2.348+07 1.178+D7 __ OfE~t06 _23 33 6 5 !~ -t _ +H
_ 413.2506946 0.70 51.236 . 85f- 96-_7.6117 J 1.728+D7 '1.128+07 ~_~ _ 9.008+O6 98 29 3 3 ! j~l ~~ I 2 ~ ~2H 767.6709945 0.33 9.007 : 851+ 149.0233 I-4.878+C8 ?-498-308 ~ 149.0233 ~-'_~ 2.678+07 8 5 ! 3 1 ~~ ; !-9 +H _ 749.0233204 0.00 I _ _. ' ~.~.!
--: .~ .. _ __._'~'226__.._:._-S~=___._~_459.2352 _I 1.658+07 8.238+06 -j-_ -.li.:~~'E-JO6 _- 22131j 5 ~y..t-!.._T._ ' !'1 °H 459.2350446 D.3i k , 8.017 .. ' APCIa __ 1 319.2267 1~3.59E+D7i 3.598+07 "___j 399.2267 1 4.488+06 _-~'P_-j 311 3 I ~, I ..~, _t __ __ +H 399.2267793 -0.22 ;_.___7.742 ' ~_ - 851._ _-249.1494-' 2.t4c+97;2.?4E+C7 _-_1._.249.9494_ j 2.778+06 15_:21j.-_.3 ... . I ; 1 -H .249.1496989 -0.77 -I-_-r-_ _ ~__..._. x.279 ~. _ I_- ESI~ . --~ 3332979 _ 9.638 07' 1 -1 Es07 .-~_ 333.207 1 ~ 9.96~+p5 -~. 20', 29 1 __ _ ø ;... . _i ! _._ ~H ! 333.297133 ~0.73 ._ x.763 -- , ..8S1+._.--1 483.7475 ~~1.43~+07~ T :68f06 _ ---_.. 1.0Q8>06 24 26 3 ..;,... 1 .... ~ _.. - +K--_., 483.1415%62 ~0.72 _ C
6.90"< .~ ESI: _- f 347.1664 ~ 1.758+07' 1.758+07 ~ 347_'186; 1.668+D6 ~.20'27 ~ _-_~_~. ~- ---Fj ~i ~7.186397G -0.77 6.655 . APCt- _ 263.1290 6.668+G6 fi.66E+D6 _~~ S.OOE+C6 ~ 1S 19 I 4 I ~-iTj '~ -H 263.1288827 0.26 o.270 _~ P _ 347.1867 1.878+07 '9.878+07 _!_34_7_.7861 2.988+D6 20 27 I 5_ ;
j,_,;~,, I 5 ~ H 347.9863976 0.83 6.0!9 j ES!+ ~,y 345.1258 5.208+07 5.028+O6 ~ I ?.008+'p6 ~14 22 6 -~j ~; ~,1 ~-1 _!~+K _- 345.1258237 ~0.09 5.306 j ESI- _3 263.7287 5.358+O6 5.3588+O6 ~ ~ 5,008+06 ' 15 19 4 I ( 1 j 9 -H 263.9288827 -0.69 !
5.309 ESI+ -229.9047 7.068+07 5.308+O6 ~ 1.OOE+Ofi 15 17 7 , -7 +N 229,1045477 0.75 I , s X75 ESI- _ 991.5076 4.978+O6 4.978+06 __ 1.908+OG _ 12 95 2~ ' 9 ~H 1 799.7977533 ~0.89_ -4,603 _ ESI~ _ 213.1494 ; 2.328+07 2.32Er_071~!_ 293.1494 _5.038;~0_6 _12 21 I 3 ; _' j ' 1 __~' 5d __i 2'13.9496987 ~1.02 4.600 ESt- -~ 277.1463 1 4,608+DS 4.6DE+~06 I I~DClE+06 96 29 I 4 7 ~H 277-7445327 -0.84 1 ___".524 1 APCI- ~ 3333074 ~ 2.20~+071 2.298+07 ~ 333.2073 1 4.8:8+06 _ 2D
29 4 1 I. .N _-i 333.207133 0.97 _ 4.963 ESI- __7, 199.1341.~_,1198E+07 5.188;=D7-,_ 8_ 199.7347 _ 1 2.838+06 _~ ~ 15 9 3y,- _ _ 1 I 9 ' -H k 999.1339681 0.6?
3 392 -._ _ ESI~ : __27/ 7650 3.778*07 3.578+07 ~ 227 1650_-r9 3:tE~-O6 __ I
.3 ~ 23T:~ 3 _i. . . '_t _ -H _ 1 227.1652682 -9.05 _ _ J___J~~.~_ _ __., _-~- 3 731 ~ - ESI+ _312 7449 _' 6.268+O6 3.538+C6 ~~ 1.008+06 5' 227 0 ~ __ I
~_-1 +H ! 372.7441639 -0.08 ' __ __ _3.177 __ ~i APCI- _; 249. 7497 ; 9.548+07 1.548+07 __ _249. 9457 I
4.958+p6 55 '21 ~__; 3 _ ' 1 _. -H_ ~ 249.7496181 9.19 _ 2 566 _ ~I APCI- 329 2336 2,298+07 2.298-07 _ 329 2335 _~-B_9!F-O6 ~ "9~-335 ~i -__~.7 _ --H ~'~. 329.2333477 0.58 ~I
?.438 ' ESI- ~ u_4_55 179_4 ~ 2.448+C6 2.448.+06 ~--~- I t_OOU+O6 ~20~1I ~ 7 _.. . : 7 - --Fi-_ ' 495.7795876 --0.6U ;
i_..._.~.097~~_._._ESI+ ____I 285.0957--1~4.03E+OE 2.028+O6_ ~-._...._...___..i-~.COE+O6 ~-._.~50~-7716:-_.-~_;21__,.. _--5 ___~_-pli._-.265.0550624 -0.0;
vonnznents: Table '~a'i illustrates ho~~ our- tech~-.ology can be used to compare she z~~zetaboiic ~rozile of an ~orgar~ism under dificezent er~~riro~zmental conditions. ~Ieae we were able to detect and identii~ Fey molec).~ies it=_vol~,led in con~troiling tree giant's response to g~~osphate conditions. This capability aalov~s researc~~e~-s to determine rwhat effects changes in I a envizonznental c.ondition~ ~,vill Cave o~~ tlm biologicai f unctions of an org;ar~isrr~.

Table VII MS/MS Data for Selected Metabolites Observed to be Increased in the -P
Fraction Parent Fragment Loss Of:

C31H3sNs010 [H+~ C19H23N6~5[H+~ C12H12~5 651 C 19H21N604[H+] C 12H1446 +ESI *CIOH9N6~2[~+J C21H24~8 C9H7 [H+]

C31 H3sNs48IH+] C 19H23N6~5 [H+~ C 12H 123 619 C19H21N6~4[H+] C12H14~4 +ESI *CppH9N(O2[H+J C21H2406 C9H~[H+~

C26H29N6~9[H+] C 19H23N6~5 [H~ C7H6~4 569 C19H21N604[H+] C~H80;

+ESI * C loH9Ns02[H+] C l6Hzo0?

C9H~[H+]

C28H43N6~6 [H+] C 19H23N6~s [H+] C9H200 559 C 19I'~21N6~4[H+] C9H22~2 +ESI *C loH9N642[H+] C 1 aH2o4a C9HyH ]

C28H43N6~5[H+] . C19H23N6~5[H+] C9H20 543 C 19H2 tN6~4 [H+] C9H22~

+ESI *CloH9N64a[H+] ClsH2o03 C9H~ [H+]

C27H3sN6O6[H+~ C 19Hz3N60s [H+] CsH t 24 539 C t9H21N6~4[H+] C8H14O2 +ESI *CisH2iN64z[H+] *C1zH1444 C loH9Ns~2[H+] C 17H26~4 C9H7[H+]

C2bH3IN6~6[H+] Ct9H23N6~s[I"i~ C7H9 523 Cl9HztN604[H+] C~Hlop2 +ES I *C14H,~N602[H+] *Ct2H14~4 C ~ oH9N602 [H+) C 16H22~4 C9H~ [H+]

C22H23N6~6[H+] *WoH9N6~2[H+] ' *C12H14O4 +ESI

* C ~ 2H ~ s04[H+] * C i oH90s [H+] C2H60 223 C9H~03[H+] C3H80 +ESI CsHs03[H+] C4Hio0 C6Hs0[H+~ C6H~o03 *C ~ oHsOs [H+] * CsHs03 [H+] C2Ha 177 C6H50[H+] C4H402 +ESI

*CgHs03[H+] C~H502[H+] CO

149 C6Hs0[H+] C202 +ESI

Table VIII Determination of Metabolite Relations using MS/MS data Rl _ R3 R2 C l oH8N602 None C ( 2H 1404 C I OH8N6~2 C4H8 C 12H I4~4 C OH8NbO2 CSH12 C12H14~4 C I OHsNsC2 C6H6 C t 2H 1 4~4 C t oHsN6C2 C4H6d3 C 12H 144 CIOHaN642 C9Hlo~z C lzHl4~4 CIOHaN642 C9H10~4 C (2H 1404 Clol'IsNs~2 C6H6 C 12H1403 Table IX: Mass Analysis of unknown peak observed in Tobacco Flower Analysis Mass Analysis of Unknown Peak Calibration Constants:
ML 1: 108299134.679450 MI.2: -16.576817 ML3: -2029.796744 Calibration Results:

Ref. Masses Exp. Diff (ppm) Masses 124.039300 124.0392980.0187 161.092070 161.0920790.0542 303.166300 303.1662720.0919 609.280660 609.2806640.0060 962.430130 962.4302300. t 037 Observed Mass of Unknown: 595.16572 Empirical Formula Search Result: C?~H3oO~s [+H]+
Mass: 595.16575 Mass Error: 0.04 ppm Proposed Metabolite: Ct5Hto06 - Rhamnoglucoside (present in flowers of grapefruit) Comments: Figures I6 and 17 and Table IX show how our technology provides meaningful information that would otherwise not be obtained. In this example the researcher thought that he knew the primary color component in tobacco flowers (ClSHlOQ6-Glucoside) but our analysis showed that the primary color component in tobacco flowers is actually the rhamnoglucoside. This illustrates the power of being able to identify unknown components after analysis. No other technology is currently available to provide this type of analysis.

2~
Table X Illustration of the number of metabolites monitored in strawberry extracts.
Summary of Metabolites Observed from Different Extraction Methods and Ionization Conditions.
Number of Unique Metabolites Observed . 5n15n ~[ Tntai ESI + 1143 1054 540 165'7 APCI + 979 1431 615 1795 Total 3986 4480 1736 6730 Table X and Figure 18 illustrate the comprehensive nature of our invention.
Our technology allows for the comprehensive comparison of the metabolic profiles of organisms under varying environmental, genetic, and developmental conditions.

In this patent document, the word "comprising" is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. A reference to an element by the indefinite article "a" does riot exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements.
It will be apparent to one skilled in the art that modifications may be made to the illustrated embodiment without departing from the spirit and scope of the invention as hereinafter defined in the Claims.

Claims (34)

1. A Method for non-targeted complex sample analysis, comprising the steps of:
providing a known molecule database (16) containing identifying data of known molecules;
introducing a complex sample containing multiple unidentified molecules into a Fourier Transform Ion Cyclotron Mass Spectrometer (FTMS) (12) to obtain identifying and quantitation data regarding the molecules in the complex sample; comparing the collected data regarding the molecules in the complex sample with the identifying data of known molecules in order to arrive at the identification of the molecules in the sample; and the creation of a non-targeted metabolite database from all the identifying and quantitation data collected from the complex sample.

2. The method as defined in Claim 1, the complex sample being a biological sample.

3. The method as defined in Claim 1, the complex sample being a combinatorial chemistry synthesis sample.

4. The method as defined in Claim 1, the identifying data being the experimentally determined empirical formula of the parent molecule and whose theoretical mass agrees to within 1.0 ppm relative error of the experimentally measured mass.

5. The method as defined in Claim 1, the identifying data being the accurate mass of the parent molecules experimentally determined with a relative error of determination less than 1.0 ppm.

6. The method as defined in Claim 1, the identifying data being the accurate mass of the fragments of the parent molecules experimentally determined with a relative error of determination less than 5.0 ppm.

7. The method as defined in Claim 1, the identifying data being the experimentally determined empirical formula of the fragment molecules of the parent molecules and whose theoretical mass agrees to within 5.0 ppm relative error of the experimentally measured mass.

8. The method as defined in Claim 1, the quantitation data being the relative and/or absolute intensity of the parent molecule.

9. The method as defined in Claim 1, the quantitation data being the relative and/or absolute intensity of the fragment molecules.

10. The method as defined in Claim 1, the non-targeted metabolite database being organized to, permit searching for known metabolites by accurate mass (defined as measured mass with less than 1.0 ppm relative error).

11. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit searching for known metabolites by empirical formula.

12. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit identification of metabolites by the accurate mass of the parent molecule.

13. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit identification of metabolites by the empirical formula of the parent molecule.

14. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit identification of metabolites by the empirical formulas of the fragments of the parent molecule.

15. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit identification of metabolites by the accurate masses of the fragments of the parent molecule.

16. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit the comparison of two samples to each other such that the relative intensity, presence, and/or absence of each metabolite is determined.

17. The method as defined in Claim 1, the non-targeted metabolite database being organized to permit the comparison of one or more "test" samples to a "control" sample such that the intensity, presence, and/or absence of the metabolites present in the "test"
samples can be determined relative to the control sample and other test samples.

18. The method as defined in Claim 1, 16, or 17, the non-targeted metabolite database being organized to permit for the sorting, presenting and reporting of the data in ascending or descending order of the relative intensities determined.

19. The method as defined in Claim 1, 16, or 17, the non-targeted metabolite database being organized to permit for the sorting, presenting and reporting of the data according to the accurate mass of the fragments of the parent molecules.

20. The method as defined in Claim 1, 16, or 17, the non-targeted metabolite database being organized to permit for the sorting, presenting and reporting of the data according to the empirical formulas of the fragments of the parent molecules.

21. The method as defined in Claim 16, 17, 18, 19, 20, the correlation of the data contained within the non-target metabolite database from biological samples from a genetically modified "test" organism and its non genetically modified "control" organism with gene expression data from same said organisms for the purpose of determining the function of the genes affected by the genetic modification.

22. The method as defined in Claim 16, 17, 18, 19, 20, the correlation of the data contained within the non-target metabolite database from biological samples from an organism exposed to a "test" environment and a "control" environment with gene expression data from same said organism under same said conditions for the purpose of determining the function of the genes affected by the test environment.

23. The method as defined in Claim 22, the test environment is deemed to be any internal or external force imparted on the organism that may have an impact on its function.
Examples include but are not limited to: exposure to or withdrawl from drug, pesticide, nutrient, or other chemical entity, weather conditions such as drought, frost, heat, psychological conditions such as stress.

24. The method as defined in Claim 16, I7, 18, 19, 20, the correlation of the data contained within the non-target metabolite database from biological samples from an organism at different stages of its development with gene expression data from same said organism at same said stages of its development for the purpose of determining the function of the genes affected by. the changes in development of the organism.

25. The method as defined in Claim 16, 17, .18, 19, 20; the correlation of the data contained within the non-target metabolite database from biological samples from a healthy organism and diseased organism with gene expression data from same said , organisms for the purpose of determining the function of the genes affected by the disease state of the organism.

26. A complex sample analysis system comprising, (a) a sample injection port for injection of a complex sample;
(b) a Fourier Transform Ion Cyclotron Mass Spectrometer (FTMS) for receiving the complex sample and separating the sample into a plurality of individual components and generating a data set for each of said components;
(c) a data processing system for transferring said data set to a first database and comparing one or more of said data sets to a second database comprising a plurality of known data sets.

27. The system of claim 26, wherein the sample injection port is automated.

28. The system of claim 26 further comprising one or more chromatographic columns connected in series to said sample injection port.

29. The system of claim 26 comprising a syringe pump.

30. The system of claim 26 wherein said data set comprises at least two characteristics of a component selected from the group consisting of mass, retention time, relative intensity, absolute intensity and combinations thereof.

31. The system of claim 26 wherein said second database comprises a plurality of data sets for known metabolites.

32. The system of claim 26 wherein said data processing system comprises a filtering system to subdivide said data set into isotopically pure subsets based upon specific mass and intensity rules for carbon, chlorine, bromine, and sulfur isotopes.

33. The system of claim 26 wherein said data processing system comprises an analytic system whereby the empirical formula and error of determination of each of the components from the subsets described in claim 32 are determined.

34. The system of claim 26 wherein said first database comprises at least two characteristics of a component selected from the group consisting of mass, retention time, relative intensity, absolute intensity, empirical formula and combinations thereof.