Method for determining small molecule components of a complex mixture and related apparatus and computer program products - Patents.com
The automated method using MS scans and symbolic algebra for chromatography data analysis addresses the inefficiencies of current techniques by forming component clusters and identifying unnamed compounds in complex mixtures, enhancing speed and accuracy.
Patent Information
- Application Number
- JP2023510333
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2020-11-20
- Filing Date
- 2021-08-13
- Publication Date
- 2024-11-11
- Estimated Expiration
- 2041-08-13
AI Technical Summary
Current methods for identifying unnamed small molecule compounds in complex mixtures are time-consuming, prone to false positives, and require significant human intervention, with existing chromatography/mass spectrometry techniques struggling to distinguish new compounds from known ones and background noise.
A fully automated method using MS scans and symbolic algebra to analyze data from liquid or gas chromatography, forming component clusters and discarding known features, focusing on candidate compounds with statistical significance, and correlating them with unknown compounds in a database.
This approach significantly reduces analysis time, enhances accuracy, and efficiently identifies new compounds by automating the process, reducing false positives and background noise, enabling faster discovery of unnamed compounds.
Smart Images

Figure 0007580577000010 
Figure 0007580577000011 
Figure 0007580577000012
Abstract
Description
[Technical field]
[0001] Aspects of the present disclosure relate to the analysis and determination of small molecule components of complex mixtures, and more particularly to methods and associated apparatus and computer program products for analyzing and identifying small molecule components or compounds of complex mixtures. Such small molecule analysis includes metabolomics, which is the study of small molecules produced by metabolic processes in living organisms, or other analysis of small molecules produced through metabolism. [Background technology]
[0002] Small molecules or compounds are extremely diverse and numerous. The total number of compounds present in nature is unknown, but it is generally estimated to be at least tens of thousands. A particular compound data repository (library or database) can store approximately 5,000 named compounds and 2,000 unnamed compounds, and therefore thousands of potentially important unnamed compounds remain to be reported and cataloged. Discoverable unnamed compounds are detectable by existing chromatographic / mass spectrometry methods in the discovery process, but such unnamed compounds are not currently reported because they do not match existing known compounds in databases or libraries.
[0003] Currently, the discovery of unnamed compounds is generally performed with the aid of graphical interactive tools. However, while such graphical interactive tools provide essential assistance to scientists, they are not fully automated, are very time consuming, find many false positives, and have other technical limitations. As a result, they are rarely used, and as a result, the number of unnamed compounds reported in samples has stagnated. There is little incentive to add unnamed compounds to a database or library unless their structure and identity are subsequently determined.
[0004] Chromatography / mass spectrometry has the ability to generate data that leads to the detection and identification of individual compounds in a sample. One way to accurately identify compounds detected in a sample is by using a library or database of trusted chemical standards. However, in many situations, it is desirable to discover unnamed novel compounds that are not already stored in such libraries or databases. The challenge in identifying unnamed novel compounds using data derived using mass spectrometry is that unnamed novel compounds are detected as ion features or characteristics (defined by characteristic mass-to-charge ratios and retention indices). In a typical sample, there are numerous features that are detected, most of which do not correspond to unique compounds but rather are overlapping representations of other known compounds. Thus, processing and characterizing the features to provide enrichment for unnamed true novel compounds can be time consuming and error prone.
[0005] In addition to identifying recurring unnamed compounds without human input, a methodology would be desirable to provide additional information about the unnamed compounds to assist scientists in prioritizing them. Additionally, this methodology would desirably remove / exclude feature data from the analysis that pertains to existing known compounds present in libraries or databases, as well as other irrelevant features such as background compound contaminants and chromatographic artifacts. A methodology would also be desirable to analyze the remaining features and output candidate compounds by consensus m / z, retention index (e.g., retention time of feature / compound normalized to retention time of adjacent eluting known ions / compounds), isotopic signature, consensus MS / MS spectrum, and peak area for optional statistical analysis. Summary of the Invention [Means for solving the problem]
[0006] The above and other needs are met by aspects of the present disclosure, which in some embodiments provide a fully automated method, apparatus, and computer program product for detecting new biochemicals in one or more samples in an automated manner that is faster and more accurate than existing methods. For example, using a commercially available desktop computer, a small scale test can be analyzed within an hour, and a study of 6200 samples was analyzed in about 8 hours. In contrast, currently commercially available software that performs a similar function of identifying biochemicals in samples is estimated to take 4-5 hours to analyze a set of 60 samples, while performing substantially fewer functions. Aspects of the present disclosure can also more effectively remove noise and other irrelevant features, and optionally remove features that have no statistical significance to the study.
[0007] In some embodiments, this methodology relies primarily on MS scans rather than integrated peaks, making it much faster and perhaps more accurate than other prior art methods. This methodology also discards features that correspond to known features associated with existing library entries (including in-source fragments, adducts, and dimers), and also discards features that do not have a reasonable molecular formula. An embodiment of the present disclosure efficiently analyzes about 1500 possible adducts and isotopic relationships among the ions in an MS scan. This analysis of a large number of mass relationships can be achieved by using software that symbolically solves mathematical equations in a separate program that itself writes part of the program code. The methodology of the present disclosure can be applied to a single sample or to multiple samples. The methodology of the present disclosure can also optionally discard features that do not have statistical significance with respect to the metadata of the study under consideration.
[0008] One particular aspect of the present disclosure provides a method for analyzing data of one or more samples, where data of each sample is obtained from a component separation and tandem mass spectrometer system comprising a separation section including a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step or function (MS1), and a second mass analysis step or function (MS2). The data from MS1 includes MS1 sample components, and the data from MS2 includes MS2 sample components. Such a method includes analyzing a data set of a plurality of MS2 sample components for each sample to determine a precursor ion mass-to-charge ratio (m / z) and retention index (RI) of each MS2 sample component. The precursor ion m / z and RI of each MS2 sample component are compared to the precursor ion mass-to-charge ratio and retention index of any MS2 sample components corresponding to one of the known compounds and the known compounds deleted from the data set. The MS2 sample components remaining in the data set are candidate MS2 sample components. Forming component clusters over the candidate MS2 sample components, each including candidate MS2 sample components having a precursor ion m / z and RI within the respective ranges of the component cluster. For each sample in each component cluster, acquiring one or more MS1 sample components within the respective precursor ion m / z and RI ranges of that component cluster. For each component cluster, determining at most one consensus MS1 sample component by aggregating one or more MS1 sample components within the respective precursor ion m / z and RI ranges, a consensus MS1 sample component representing each sample. Associating the consensus MS1 sample component with a corresponding consensus MS2 sample component, consensus precursor ion m / z, and consensus RI determined by aggregating the MS2 sample components and associated precursor ion m / z and RI in the component cluster. For each component cluster, determining whether the consensus MS1 sample component exhibits a molecular ion or a derivative relationship to a molecular ion.The component clusters are grouped according to consensus RI, and from each group of component clusters, one or more component clusters are selected that are candidate component clusters. The consensus precursor ion m / z, consensus RI, consensus MS1 sample components, and consensus MS2 sample components of the candidate component clusters are then correlated with unknown compounds in the sample.
[0009] Thus, the present disclosure includes, but is not limited to, the following exemplary embodiments:
[0010] Exemplary embodiment 1: A method of analyzing data for one or more samples, the data for each sample being obtained from a component separation and tandem mass spectrometer system comprising a separation section comprising a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step or function (MS1), and a second mass analysis step or function (MS2), the data from MS1 comprising MS1 sample components and the data from MS2 comprising MS2 sample components, the method comprising: for each sample, analyzing a dataset of a plurality of MS2 sample components to determine a precursor ion mass to charge ratio (m / z) and retention index (RI) of each MS2 sample component; comparing the precursor ion m / z and RI of each MS2 sample component to the precursor ion mass to charge ratio and retention index of known compounds and removing from the dataset any MS2 sample component that corresponds to one of the known compounds, wherein the MS2 sample components remaining in the dataset are candidate MS2 sample components. forming component clusters over the candidate MS2 sample components including candidate MS2 sample components each having a precursor ion m / z and RI within the range of the component cluster; for each sample in each component cluster, obtaining one or more MS1 sample components within the precursor ion m / z and RI range of that component cluster; for each component cluster, determining at most one consensus MS1 sample component to represent each sample by aggregating one or more MS1 sample components within the respective precursor ion m / z and RI range, and associating the consensus MS1 sample component with a corresponding consensus MS2 sample component, consensus precursor ion m / z, and consensus RI determined by aggregating the MS2 sample components and associated precursor ion m / z and RI within the component cluster; for each component cluster, determining whether the consensus MS1 sample component exhibits a molecular ion or a derivative relationship to a molecular ion;grouping the component clusters according to consensus RI and selecting one or more component clusters from each group of component clusters that are candidate component clusters; correlating the consensus precursor ion m / z, consensus RI, consensus MS1 sample components, and consensus MS2 sample components of the candidate component clusters with unknown compounds in the sample;
[0011] Exemplary embodiment 2: A method or combination thereof according to any of the preceding exemplary embodiments, wherein removing any MS2 sample components corresponding to one of the known compounds from the dataset comprises removing any of a known molecular ion corresponding to one of the known compounds, background artifacts, chromatographic artifacts, or derivative relationships of a known molecular ion, mass relationships of a known molecular ion, background artifacts, or chromatographic artifacts from the dataset.
[0012] Exemplary embodiment 3: A method or combination thereof according to any of the preceding exemplary embodiments, wherein removing derivative relationships of known molecular ions or mass relationships of known molecular ions from the dataset comprises removing any of in-source fragments, adducts, isotopes, dimers, oligomers, multiply charged species, isotopic adducts, or isotopic oligomers from the dataset.
[0013] Exemplary embodiment 4: A method according to any of the preceding exemplary embodiments or combinations thereof, comprising determining a precursor ion mass to charge ratio and retention index of a known compound from an MS1 sample component of the known compound.
[0014] Exemplary embodiment 5: A method or combination thereof according to any of the preceding exemplary embodiments, comprising associating any of the precursor ion m / z, RI, MS1 sample components, and MS2 sample components of each known compound with the known compounds in a database comprising an ion data repository.
[0015] Exemplary embodiment 6: A method or combination thereof according to any of the preceding exemplary embodiments, comprising associating the precursor ion m / z, RI, MS1 sample component, and MS2 sample component of each candidate component cluster with an unknown compound in a database comprising an ion data repository.
[0016] Exemplary embodiment 7: A method according to any of the preceding exemplary embodiments or a combination thereof, comprising determining a retention index by normalizing the retention time of each MS2 sample component.
[0017] Exemplary embodiment 8: A method according to any of the preceding exemplary embodiments or a combination thereof, wherein forming component clusters of the candidate MS2 sample components comprises forming component clusters across the candidate MS2 sample components using partitional clustering or agglomerative clustering.
[0018] Exemplary embodiment 9: A method according to any of the preceding exemplary embodiments or combinations thereof, comprising sorting the candidate MS2 sample components by precursor ion mass before forming component clusters across the candidate MS2 sample components.
[0019] Exemplary embodiment 10: A method according to any of the preceding exemplary embodiments or combinations thereof, comprising sorting the component clusters by retention index to generate secondary component clusters.
[0020] Exemplary embodiment 11: A method or combination thereof according to any of the preceding exemplary embodiments, wherein the step of forming component clusters across the candidate MS2 sample components is repeated two or more times.
[0021] Exemplary embodiment 12: A method or combination thereof according to any of the preceding exemplary embodiments, wherein comparing the consensus MS1 sample components of each component cluster across component clusters is performed across multiple samples.
[0022] Exemplary embodiment 13: A method according to any of the preceding exemplary embodiments or combinations thereof, comprising removing any component cluster that is present in less than 5% of the plurality of samples.
[0023] Exemplary embodiment 14: A method or combination thereof according to any of the preceding exemplary embodiments, comprising analyzing multiple possible derivative relationships or possible mass relationships of MS1 sample components prior to associating the consensus MS1 sample components with the corresponding consensus precursor ion m / z.
[0024] Exemplary embodiment 15: A method according to any of the preceding exemplary embodiments or combinations thereof, wherein analyzing the plurality of possible derivative relationships or the plurality of possible mass relationships includes creating computer code for analyzing the plurality of possible derivative relationships or the possible mass relationships in one run using symbolic algebra to analyze a specified set of derivative relationships or mass relationships.
[0025] Exemplary embodiment 16: A method or combination thereof according to any of the preceding exemplary embodiments, including removing any component clusters that do not have a valid molecular formula prior to grouping the component clusters.
[0026] Exemplary embodiment 17: A method according to any of the preceding exemplary embodiments or combinations thereof, comprising removing any component clusters that do not have statistical significance from the metadata of the data of the one or more samples being analyzed prior to grouping the component clusters.
[0027] Exemplary embodiment 18: A method or combination thereof according to any of the preceding exemplary embodiments, comprising prioritizing candidate component clusters having statistical significance as candidate unknown compounds with respect to metadata of data of one or more samples analyzed.
[0028] Exemplary embodiment 19: An apparatus for analyzing data of one or more samples, comprising: Data for each sample is obtained from a component separation and tandem mass spectrometer system comprising a separation section comprising a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step or function (MS1), and a second mass analysis step or function (MS2); A processor and a memory storing executable instructions that, when executed by the processor, cause an apparatus to perform at least the steps or a combination thereof of the method according to any of the exemplary embodiments described above; An apparatus comprising:
[0029] Exemplary embodiment 20: A computer program product for analyzing data of one or more samples, comprising: Data for each sample is obtained from a component separation and tandem mass spectrometer system comprising a separation section including a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step (MS1), and a second mass analysis step (MS2); A computer program product comprising at least one non-transitory computer readable storage medium having computer readable program code stored thereon, the computer readable program code comprising program code for performing the steps or combinations of the methods according to any of the preceding exemplary embodiments.
[0030] These and other features, aspects, and advantages of the present disclosure will become apparent from a reading of the following detailed description in conjunction with the accompanying drawings, which are briefly described below. The present disclosure includes any combination of two, three, four, or more features or elements described in the present disclosure, regardless of whether such features or elements are explicitly combined or recited in the description of a particular embodiment herein. The present disclosure, in any of its aspects and embodiments, is intended to be read holistically such that any separable features or elements of the disclosure are deemed combinable as intended, i.e., unless the context of the disclosure clearly dictates otherwise.
[0031] It will be understood that the summary of this specification is provided only for the purpose of summarizing some exemplary aspects to provide a basic understanding of the present disclosure. Therefore, it will be understood that the exemplary aspects described above are merely examples and should not be construed to narrow the scope or spirit of the present disclosure in any way. It will be understood that the scope of the present disclosure encompasses many potential aspects in addition to those summarized herein, some of which are further described below. Furthermore, other aspects disclosed herein and advantages of such aspects will become apparent from the following detailed description taken in conjunction with the accompanying drawings, which illustrate by way of example the principles of the described aspects.
[0032] Having thus described the disclosure in general terms, reference is now made to the accompanying drawings, which are not necessarily drawn to scale. [Brief description of the drawings]
[0033] [Figure 1] FIG. 1 illustrates a schematic diagram of a system according to one embodiment of the present disclosure, including a memory device having a database, a processor device, and a user interface (display) in communication with a spectroscopic analysis device. [Diagram 2] FIG. 1 is a schematic diagram illustrating a three-dimensional plot of spectroscopic data associated with one exemplary sample. [Diagram 3] FIG. 2A is a schematic diagram illustrating a two-dimensional profile plot of one exemplary sample, which may be determined from a corresponding three-dimensional plot of spectroscopic data for that sample, according to some embodiments of the present disclosure. [Figure 4] FIG. 2A is a schematic diagram illustrating a two-dimensional profile plot of one exemplary sample, which may be determined from a corresponding three-dimensional plot of spectroscopic data for that sample, according to some embodiments of the present disclosure. [Diagram 5] FIG. 1 shows a schematic diagram of a method for analyzing and identifying small molecule components or compounds of a complex mixture according to one exemplary embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0034] The present disclosure will now be described more fully hereinafter with reference to the accompanying drawings, in which some, but not all, aspects of the disclosure are shown. Indeed, the disclosure may be embodied in many different forms and should not be construed as limited to the aspects set forth herein, but rather, these aspects are provided so that this disclosure will satisfy applicable legal requirements. Like numbers refer to like elements throughout.
[0035] Definition of Terms The terms "compound", "small molecule", "metabolite", or "biochemical" may be used interchangeably and refer to organic and inorganic molecules present in a cell. This term does not include large macromolecules such as large proteins (e.g., proteins with molecular weights greater than 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, or 10,000), large nucleic acids (e.g., nucleic acids with molecular weights greater than 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, or 10,000), or large polysaccharides (e.g., polysaccharides with molecular weights greater than 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, or 10,000). Small molecules of cells are generally found free in solution in the cytoplasm or other organelles such as mitochondria, where they form a pool of intermediates that can be further metabolized or used to generate larger molecules called macromolecules. The term "small molecule" includes signaling molecules and intermediates in chemical reactions that convert energy from food into usable forms. Non-limiting examples of small molecules include sugars, fatty acids, amino acids, nucleotides, intermediates formed during cellular processes, and other small molecules found within cells.
[0036] The term "retention index" or "RI" is a normalized measure of a compound's retention time in liquid chromatography.
[0037] The term "unnamed compound" or "unnamed biochemical" refers to a compound that is recognized by mass, RI, and MS2 spectra, but whose chemical name, molecular formula, and chemical structure are unknown.
[0038] The term "tandem MS" refers to an operation in which a first MS step, called "primary MS" or "MS1", is performed followed by one or more subsequent MS steps, commonly called "secondary MS" or "MS2". In the primary MS (MS1), ions representing one (or sometimes more than one) chemical component (MS1 sample component) are detected and recorded in producing a primary mass spectrum. The material represented by the ions is subjected to secondary MS (MS2) in which the material of interest is fragmented to break the material into subcomponents, which are detected and recorded as secondary mass spectra (MS2 sample components). In true tandem MS, there is a clear relationship between the ions of interest in the primary MS and the resulting peaks produced during the secondary MS. The ions of interest in the primary MS correspond to "parent" or precursor ions, while the ions produced during the secondary MS correspond to subcomponents of the parent ions, and are referred to herein as "daughter", "child" or "product" ions. Tandem MS allows the creation of data structures that represent the parent-daughter relationships of chemical components in a complex mixture. This relationship can be represented by a tree-like structure that shows the mutual relationships of parent and daughter ions, where the daughter ions represent subcomponents of the parent ion. For example, tandem MS can be repeated on daughter ions to determine "grand-daughter" ions. Thus, tandem MS is not limited to two levels of fragmentation, but can be used to perform multiple "MS n "The term "MS / MS" is commonly used to refer to multilevel MS, also known as "MS2." The term "MS / MS" is a synonym for "MS2."
[0039] The term "ionization" is the process by which neutral molecules become charged ions by the addition or removal of charged subatomic particles (e.g., protons or electrons) that enable MS detection. "Ionization products" are ions formed by a single molecule or compound in the ionization process.
[0040] The term "MS1 spectrum" or "primary mass spectrum" or "MS1 sample component" or "MS1 SC" refers to sample data obtained from primary mass spectrometry (MS1).
[0041] The terms "MS2 spectrum" or "secondary mass spectrum" or "MS2 sample components" or "MS2 SC" refer to sample data obtained from secondary mass spectrometry (MS2).
[0042] The term "precursor ion mass" refers to the mass of the ion of interest detected in the primary MS (MS1) step.
[0043] The term "library" or "ion data repository" refers to a collection of information about compounds detected by mass spectrometry. The information about the compounds may include, for example, information related to the mass, RI, and MS spectrum of the compounds. The information may also include information related to various isotopes and adducts of the compounds. The library may also include information from public databases such as SMILES strings, Inchi strings, InchiKey, etc.
[0044] The term "derivative relationship" or "mass relationship" refers to a derivative of a molecule of interest that is related to the ionization products of the molecule of interest. Derivative relationships or mass relationships can include in-source fragments, adducts, isotopes, dimers, oligomers, multiply charged species. Derivative relationships or mass relationships can also include more complex relationships, including variations of the described derivative relationships or mass relationships, including, for example, isotopic adducts, isotopic oligomers, different isotopes of the same molecule.
[0045] The various aspects of the present disclosure described above, as well as many other aspects of the present disclosure, are described in further detail herein. Apparatus, methods, and computer program products related to aspects of the present disclosure are illustratively disclosed in conjunction with an appropriate analysis device, which may include a separator or separation section (e.g., a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer) and / or a detector section (e.g., a spectrometer). However, those skilled in the art will understand that such disclosure is for illustrative purposes only to describe the implementation of various aspects of the present disclosure.
[0046] In particular, the apparatus, methods, and computer program products associated with aspects of the present disclosure can be adapted to any number of processes used to generate complex sets of data for each sample (e.g., within a single sample) or across multiple samples, whether biological, chemical, or biochemical in nature. For example, aspects of the present disclosure can be used and applied with a variety of different analytical devices and processes, including, but not limited to, analytical devices including separator sections (or "component separators" or "component separation" sections) including liquid chromatographs (LC), gas chromatographs (GC), supercritical fluid chromatographs (SFC), capillary electrophoresis (CE) analyzers, associated detector sections (or "mass spectrometer" sections) including nuclear magnetic resonance imaging (NMR) devices, mass spectrometers (MS), ion mobility mass spectrometers (IMS-MS), and electrochemical arrays (EC), and / or combinations thereof (e.g., tandem mass spectrometers including MS1 and MS2).
[0047] In some aspects of the present disclosure, the detector section may be used without the separator section. In this regard, one skilled in the art will appreciate that the aspects of the present disclosure disclosed herein are not limited to metabolomics analysis. For example, the aspects of the present disclosure disclosed herein may be implemented in other applications that require characterization or analysis of small molecules present in a sample or complex mixture, regardless of the origin of the sample or complex mixture. For example, the aspects of the present disclosure disclosed herein may also be implemented in bioprocess optimization procedures that aim to grow cells to produce drugs or additives, or in drug metabolite profiling procedures that aim to identify all metabolites that are the result of the biotransformation of an administered xenobiotic. Some other non-limiting examples of other applications may include quality assurance procedures for the manufacture of consumer products, where the objective may be to objectively ensure that desired product characteristics are met in procedures where a large number of sample components can produce a particular attribute, such as taste or flavor (e.g., cheese, wine, or beer) or aroma / odor (e.g., fragrance). Thus, one common theme exhibited by the embodiments of the present disclosure disclosed herein is that small molecules in a sample can be analyzed using the various apparatus, method, and computer program product embodiments disclosed herein.
[0048] FIG. 1 shows an example of a system according to one embodiment of the present disclosure, which is in communication with an analytical device 110, such as a combination chromatograph (component separator / component separation) / tandem mass spectrometer (MS1, MS2). However, those skilled in the art will recognize that the configuration of the analytical device 110 presented herein is for illustrative purposes only and is not intended to be limiting with respect to the range of suitable and appropriate analytical devices that may also be applied under the principles disclosed herein. As shown, a sample 100 (whether biological, chemical, or biochemical in nature) is introduced into the separator / separation portion of the analytical device 110 and, when applied through the first mass spectrometer (MS1) and second mass spectrometer (MS2) of the detector portion, may be analyzed using suitable techniques recognized by those skilled in the art.
[0049] For example, components of a particular sample 100 may pass through a column associated with the separator section / separator section at different rates and exhibit different spectral responses (e.g., related to intensity as a function of retention time) as detected by a first mass analysis step (MS1) of the detector section based on their specific characteristics. A second mass analysis step (MS2) adds a second phase of mass fragmentation that may be performed, for example, to facilitate quantification of low level compounds in the presence of a high sample matrix background. As will be appreciated by those skilled in the art, the analysis device 110 can generate a set of spectroscopic analysis data corresponding to each sample 100 and having three or more dimensions (e.g., quantifiable sample properties) associated therewith, and the data included in the data set generally indicates the composition of the sample 100 (e.g., sample components). In some embodiments, the data set may include data for each sample relating, for example, to retention time, mass (or mass-to-charge ratio) of the sample or components (ions), intensity, or even an index or identity of the sample. However, such data must first be properly analyzed to determine the composition of the sample (e.g., ions, metabolites).
[0050] In some cases, a three-dimensional data set (MS1 or MS2) for each of one or more samples, with each dimension corresponding to a quantifiable sample characteristic, may be selected or designated for further analysis. An example of such a three-dimensional set of spectroscopic data is generally shown in FIG. 2 and may be plotted on a three-axis plot or graph including individual axes for response intensity element 220, sample component mass element 210 (or mass-to-charge ratio), and time element 230 (specifically, in this example, retention time, or the time a particular component spends in the column of the separator portion of the analytical device 110). That is, the data obtained for a particular sample includes, in some embodiments, the relationship between ion mass 210, retention time 230, and intensity 220, including intensity 220 as a function of retention time 230 of a particular ion mass 210. The location of the data points relative to the sample component mass axis 210 may indicate, for example, the number of individual component molecules in the sample 100 and the relative mass values of such sample components.
[0051] According to other aspects of the present disclosure, various analytical devices may be used to generate three or more dimensional sets of analytical data corresponding to the sample 100. For example, the analytical devices may include various combinations of separator / separator sections including one of a liquid chromatograph (LC) (positive or negative channel) and a gas chromatograph (GC), a supercritical fluid chromatograph (SFC), a capillary electrophoresis (CE) analyzer, and associated detector sections including, but not limited to, one of a nuclear magnetic resonance imaging (NMR) device, a mass spectrometer (MS), an ion mobility spectrometer (IMS), a tandem mass spectrometer (MS1 and MS2), and an electrochemical array (EC). In some aspects, the analytical device may include a detector section without a separator section. Those skilled in the art will recognize that such complex three or more dimensional data sets may be generated by other suitable analytical devices that may communicate with the components of the aspects of the present disclosure as described in further detail herein.
[0052] One or more samples 100 may be individually withdrawn from well plates 120 and / or other types of sample containers and individually introduced to the analytical device 110 for analysis and formation of a corresponding three or more dimensional data set (see, e.g., FIG. 2). For example, individual samples 100 may be transferred from the well plates 120 to the analytical device 110 via pipettes, syringes, microfluidic channels defined by a test array, and / or other systems for transferring samples in a laboratory environment. As disclosed herein, the nature of the samples is quite diverse and generally includes mixtures or complex mixtures containing small molecules, which may illustratively include, but are not limited to, blood samples, urine samples, cell cultures, saliva samples, plant tissues and organs (e.g., leaves, roots, stems, flowers, etc.), plant extracts, culture media, membranes, cell compartments / organelles, cerebrospinal fluid (CSF), milk, soda products, food products (e.g., yogurt, chocolate, juice), and / or other types of biological, chemical, and / or biochemical samples in which metabolites and / or chemical / molecular components of interest may be present. One common aspect of these possible samples or sample types is that the selected sample includes MS1 and / or MS2 sample components. Empirical data or other information related to the MS1 and / or MS2 sample components of a sample may be implemented to determine, for example, one or more ions, small molecules, or metabolites expected to be present in a sample having such MS1 and / or MS2 sample components. That is, such information relating to the MS1 and / or MS2 sample components provides a context to the sample and the data obtained from the sample via the component separation and mass spectrometer system, and the context provides an indication or indications regarding at least the base components or building blocks of the sample.
[0053] As shown in FIG. 1, aspects of the disclosure may include an ion data repository (e.g., a library) including, for example, a database (e.g., a relational database) that is at least partially stored, for example, as executable or accessible instructions in a memory or memory device 140 (i.e., a computer readable storage medium on which computer readable program code portions are stored), which communicates with a processor or processor device 130 (e.g., a computing device implementing a processor) to selectively execute the instructions / computer readable program code portions in the memory device 140 to cause the apparatus to perform certain method steps and / or functions. In some cases, the memory device 140 and / or the processor device 130 may be configured to directly or indirectly communicate with the analytical device 110 to receive a dataset corresponding to the sample 100 from the analytical device 110 (in some cases, a dataset including three or more dimensions, where data parameters such as signature of the sample, mass of the sample or components (ions), retention time, and intensity / response may represent any one of the dimensions of the dataset). That is, the data set received by the memory device includes, for example, data indicative of a relationship between ion mass (or mass-to-charge ratio of ions), retention time, and intensity. In some particular examples, the data set (for each of one or more samples 100) includes data indicative of intensity as a function of retention time for particular ion masses. The processor device 130 may communicate with the analytical device 110 via wired (RS-232 and / or other types of wired connection) and / or wireless (e.g., RF, IR, or other wireless communication, etc.) techniques such that a database associated with (and / or in communication with) the memory device 140 / processor device 130 can receive and store the data set from the analytical device 110.
[0054] Additionally, the analytical device 110 may communicate with one or more processor devices 130 (and associated user interfaces and / or displays 150) via wired and / or wireless computer networks, including but not limited to the Internet, local area networks (LANs), wide area networks (WANs), or other networking types and / or technologies recognized by those skilled in the art. The user interface / display 150 may be used to receive user input and communicate output, such as, for example, displaying any or all of the system-related communications, including the manipulation and analysis of sample data as disclosed herein, as will be understood and appreciated by those skilled in the art. The database may be structured using commercially available software, such as, for example, Oracle, Sybase, DB2, or other database software. As shown in FIG. 1, the processor device 130 may communicate with the user interface / display 150 and memory device 140 (such as a hard drive, memory chip, flash memory, RAM module, ROM module, and / or other memory device 140) to automatically (directly) or indirectly store / manage an ion data repository / database that includes the data sets received from the analytical device 110. Additionally, the memory device 140 may also be used to store other received data or information relating to the sample or its components in an ion data repository / database and / or data otherwise manipulated by the processor device 130.
[0055] The processor device 130 may, in some embodiments, be capable of converting a data set (see, e.g., FIG. 2, where an exemplary data set is a three-dimensional data set) containing data indicating relationships between various sample parameters, such as, for example, ion mass, retention time, and intensity, for each sample received by the memory device 140 into at least one corresponding two-dimensional data set (see, e.g., FIG. 3). The at least one two-dimensional data set may include, for example, a two-dimensional component “profile” of a particular sample 100 at a particular point 235 (FIG. 2) along one of the three axes of the three-dimensional data set. The particular point 235 along one of the three axes may be, for example, the mass of a particular selected sample component along the sample component mass axis 210. Once that particular sample component mass is selected, the resulting “slice” of the three-dimensional data set becomes a two-dimensional profile plot of the sample. That is, the resulting profile (also referred to herein as a "profile plot" as shown in Figures 3 and 4) shows the mass (and its detection intensity) of that particular sample component detected as a function of time measured from a zero point, which corresponds to the time of injection and / or other introduction of the sample 100 into the analytical device 110). For example, the processor device 130 may be configured to form a two-dimensional profile of the detection intensity / response of the sample versus / as a function of the sample component (retention) time of the sample for that given or selected sample component mass point 235 (see, e.g., Figures 3 and 4). The "x" axis (or e.g. (retention) time axis 230) of Figure 2 may further be characterized as a retention index (e.g., retention time of an ion / compound normalized to the retention time of a nearby eluting known ion / compound) and / or retention time, as the case may be.Thus, the processor device 130 may further parse the three-dimensional (or more) data set for each of the plurality of samples into one or more individual two-dimensional profiles (i.e., intensity / response profile versus retention time of the sample component) corresponding to at least one particular (selected) sample component mass point (e.g., element 235) to convert each three-dimensional (or more) data set (e.g., of FIG. 2) into at least one corresponding two-dimensional data set (e.g., having a profile or profile plot as shown in FIGS. 3 and 4) of the selected sample component, which may be further plotted as response intensity 220 of the corresponding sample component mass versus retention time 230 (or retention index) of the sample component and optionally displayed on the user interface / display 150. Those skilled in the art will recognize that any amount of two-dimensional data sets or profile plots may be formed or obtained from any three-dimensional or more data set by selecting two different sample parameters at a selected particular value of a third sample parameter and then plotting the two different sample parameters against each other in a two-dimensional plot.
[0056] According to some embodiments, the processor device 130 may be configured to selectively execute executable instructions / computer readable program code portions stored in the memory device 140, in conjunction with an ion data repository / library / database also stored in the memory device 140 as necessary, to achieve, for example, identification, quantification, display, curation, and / or other analysis of selected sample components (i.e., metabolites, molecules, or ions, or portions thereof) in each of the multiple samples (or in a single sample) from a two-dimensional data set representing each sample of the multiple samples.
[0057] According to certain aspects, as shown in FIG. 5, a method is provided for analyzing data for one or more samples (block 500), where the data for each sample is obtained from a component separation and tandem mass spectrometer system comprising a separation section (block 505) comprising a liquid chromatograph, gas chromatograph, supercritical fluid chromatograph, or capillary electrophoresis analyzer, a first mass analysis step or function (MS1) (block 510), and a second mass analysis step or function (MS2) (block 515), where the data from MS1 comprises MS1 sample components (primary mass spectrum, block 520) and the data from MS2 comprises MS2 sample components (secondary mass spectrum, block 525).
[0058] Such a method includes, for each sample, analyzing a dataset of multiple MS2 sample components to determine a precursor ion mass-to-charge ratio (m / z) and retention index (RI) for each MS2 sample component (Block 530). The precursor ion m / z and RI for each MS2 sample component are compared to the precursor ion mass-to-charge ratio and retention index of any MS2 sample components that correspond to known compounds and one of the known compounds that has been removed from the dataset (Block 535). The MS2 sample components remaining in the dataset are candidate MS2 sample components.
[0059] Component clusters are formed over the candidate MS2 sample components, each of which includes candidate MS2 sample components having a precursor ion m / z and RI within the range of the respective component cluster (block 540). For each sample in each component cluster, one or more MS1 sample components within the precursor ion m / z and RI range of the component cluster are obtained. For each component cluster, at most one consensus MS1 sample component is determined to represent each sample by aggregating one or more MS1 sample components within the precursor ion m / z and RI range (block 545). The consensus MS1 sample component is associated with the corresponding consensus MS2 sample component, consensus precursor ion m / z, and consensus RI determined by aggregating the MS2 sample components and associated precursor ion m / z and RI in the component cluster (block 550). For each component cluster, it is determined whether the consensus MS1 sample component exhibits a molecular ion or a derivative relationship to a molecular ion (block 555). The component clusters are grouped according to consensus RI, and from each group of component clusters, one or more component clusters are selected that are candidate component clusters (Block 560). The consensus precursor ion m / z, consensus RI, consensus MS1 sample components, and consensus MS2 sample components of the candidate component clusters are then correlated with unknown compounds in the sample (Block 565).
[0060] In other words, such a method includes analyzing a data set of a plurality of MS2 sample components for each sample to determine a retention index and a precursor ion mass for each MS2 sample component. The retention index and precursor ion mass of each MS2 sample component are compared to the retention index and precursor ion mass of any MS2 sample components corresponding to the known compounds and one of the known compounds deleted from the data set. The MS2 sample components remaining in the data set are candidate MS2 sample components. Such a method further includes forming component clusters over the candidate MS2 sample components, each of which includes candidate MS2 sample components having a retention index and precursor ion mass within a respective range. For each component cluster, obtain one or more MS1 sample components having a retention index and precursor ion mass within the respective retention index and precursor ion mass range. Select at most one MS1 sample component from the one or more MS1 sample components having a retention index and precursor ion mass within the respective retention index and precursor ion mass range. Such a method further includes comparing the selected at most one MS1 sample component for each component cluster across the component clusters to determine whether the selected at most one MS1 sample component is a precursor ion or a mass related to a precursor ion. If the selected at most one MS1 sample component is a precursor ion, determining a molecular mass within the range of the precursor ion masses of the component cluster and an associated retention index of the precursor ion. Forming candidate component clusters from the candidate MS2 sample components having retention indices and precursor ion masses corresponding to the determined molecular masses and associated retention indices of the precursor ions, and associating each candidate component cluster with a candidate unknown compound.
[0061] In some embodiments, removing any MS2 sample components corresponding to one of the known compounds from the dataset comprises removing any of the known molecular ions, background artifacts, chromatographic artifacts, or derivative relationships of the known molecular ions, mass relationships of the known molecular ions, background artifacts, or chromatographic artifacts from the dataset that correspond to one of the known compounds. In other embodiments, removing any of the derivative relationships of the known molecular ions or mass relationships of the known molecular ions from the dataset comprises removing any of the in-source fragments, adducts, isotopes, dimers, oligomers, multiply charged species, isotopic adducts, or isotopic oligomers from the dataset.
[0062] In some embodiments, such methods include determining precursor ion mass to charge ratios and retention indices of known compounds from MS1 sample components of known compounds.
[0063] In further embodiments, such methods include associating any of the precursor ion m / z, RI, MS1 sample component, and MS2 sample component of each known compound with the respective known compound in a database comprising an ion data repository, and / or the precursor ion m / z, RI, MS1 sample component, and MS2 sample component of each candidate component cluster with the unknown compound in a database comprising an ion data repository.
[0064] In some embodiments, such methods include determining a retention index by normalizing the retention time of each MS2 sample component.
[0065] In some embodiments, forming component clusters across the candidate MS2 sample components comprises forming component clusters across the candidate MS2 sample components using divisive clustering or agglomerative clustering.
[0066] In some embodiments, such methods include sorting the candidate MS2 sample components by precursor ion mass prior to forming component clusters of the candidate MS2 sample components.
[0067] In some aspects, such methods include sorting the component clusters by retention index to generate secondary component clusters.
[0068] In some embodiments, the step of forming component clusters across the candidate MS2 sample components is repeated two or more times.
[0069] In some embodiments, comparing the consensus MS1 sample components of each component cluster across component clusters is performed across multiple samples.
[0070] In some embodiments, such methods comprise removing component clusters that are present in less than 5% of the plurality of samples.
[0071] In some embodiments, such methods include analyzing multiple possible derivative relationships or possible mass relationships of the MS1 sample components prior to associating the consensus MS1 sample components with the corresponding consensus precursor ion m / z.
[0072] In some embodiments, analyzing the plurality of possible derivative relationships or the plurality of possible mass relationships includes generating computer code for analyzing the plurality of possible derivative relationships or the possible mass relationships in one run using symbolic algebra to analyze a specified set of derivative relationships or mass relationships.
[0073] In some embodiments, such methods include removing any component clusters without a plausible molecular formula prior to grouping the component clusters, while other embodiments include removing component clusters without statistical significance from the metadata of the data of the one or more samples being analyzed prior to grouping the component clusters.
[0074] In some embodiments, such methods include prioritizing candidate component clusters that have statistical significance with respect to metadata of data of one or more samples analyzed as candidate unknowns.
[0075] Some embodiments further include an apparatus for analyzing data of one or more samples, the data of each sample being obtained from a component separation and tandem mass spectrometer system comprising a separation section comprising a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step or function (MS1), and a second mass analysis step or function (MS2), the apparatus comprising a processor and a memory storing executable instructions that, in response to execution by the processor, cause the apparatus to perform at least the method steps according to any of the foregoing exemplary embodiments or combinations thereof disclosed herein.
[0076] Some embodiments further include a computer program product for analyzing data of one or more samples, the data of each sample being obtained from a component separation and tandem mass spectrometer system comprising a separation section including a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, a first mass analysis step (MS1), and a second mass analysis step (MS2), the computer program product including at least one non-transitory computer readable storage medium having computer readable program code stored thereon, the computer readable program code including program code for performing method steps according to any of the foregoing exemplary embodiments or combinations thereof disclosed herein.
[0077] In some embodiments, the retention index and precursor ion mass of the known compounds can be generated by MS1. In other embodiments, the retention index and precursor ion mass of the known compounds can be stored in a database. In yet other embodiments, the database can be an ion data repository.
[0078] In some aspects, the component clusters may be formed by clustering techniques such as divisive clustering or agglomerative clustering.
[0079] In some embodiments, the candidate MS2 sample components used to form the component clusters may be sorted by precursor ion mass prior to forming the component clusters. In another embodiment, the component clusters may also be sorted by retention index. The component clusters may be sorted first by precursor ion mass and then by retention index. Alternatively, the component clusters may be sorted first by retention index and then by precursor ion mass. The component clusters may be sorted iteratively and divided into progressively smaller component clusters.
[0080] In some embodiments, the process of comparing selected MS1 sample components across component clusters can be performed across multiple samples.In other embodiments, when this comparison is performed across multiple samples, component clusters that are only present in some samples can be removed.For example, component clusters that are present in less than 5% of multiple samples can be removed from the analysis.
[0081] In some embodiments, the mass relationships of the MS1 sample components may be analyzed before selecting a molecular mass corresponding to a range of precursor ion masses of the component cluster. In one example, 500 or more mass relationships may be analyzed. In another example, 1,000 or more mass relationships may be analyzed. In yet another example, 1,500 or more mass relationships may be analyzed. That is, such a method may also include analyzing a plurality of possible mass relationships of the MS1 sample components before determining the molecular mass of the precursor ion within the range of precursor ion masses of the component cluster. In this embodiment, the symbolic algebra software may be used in a computer program product to perform the analysis. In another embodiment, the symbolic algebra software may be executed once to create computer code for analyzing a set of specified mass relationships. That is, in some embodiments, analyzing a plurality of possible mass relationships includes using symbolic algebra to analyze a plurality of possible mass relationships in a single execution to create computer code for analyzing a set of specified mass relationships.
[0082] In some aspects, the component clusters may be further evaluated prior to forming the candidate component clusters. In one example of such an aspect, the component clusters may be evaluated based on molecular formula, and component clusters without a plausible molecular formula may be removed from the analysis. In another example of such an aspect, the component clusters may be evaluated based on statistical significance with respect to the metadata of the study under consideration. If the component cluster is not determined to be statistically significant based on the metadata of the study under consideration, the component cluster may be removed from the analysis. That is, in various aspects, the method may further include removing component clusters without a plausible molecular formula prior to forming the candidate component clusters, or removing component clusters without statistical significance from the metadata of the data to be analyzed prior to forming the candidate component clusters.
[0083] In some embodiments, the method may be performed on a single sample.
[0084] In some embodiments, candidate component clusters may be prioritized as candidate unknowns based on statistical significance with respect to the metadata of the study under consideration. That is, such methods may further include prioritizing candidate component clusters having statistical significance with respect to the metadata of the data being analyzed as candidate unknowns. In a feature of such an embodiment, candidate component clusters having high statistical significance may be given high priority as candidate unknowns. In another feature of such an embodiment, candidate component clusters having low statistical significance may be given lower priority as candidate unknowns.
[0085] In some embodiments, the list of candidate unknowns generated by the method as well as their associated candidate component clusters may be stored in a database for future analysis.
[0086] In some embodiments, there is information associated with the candidate unknown compound. The information associated with the candidate unknown compound may include precursor ion mass, retention index, MS1 sample components, and MS2 sample components.
[0087] Inputs and outputs: The input to the disclosed method, apparatus, and computer program product is chromatography and mass spectrometry data from one or more samples. In certain embodiments, the data is obtained from a liquid chromatograph / tandem mass spectrometer (LC-MS / MS) system. The output is a set of unnamed candidate compounds present in one or more samples that were not previously present in the ion data repository. The output also includes the mass (or mass to charge ratio m / z), retention index (or RI), and MS2 spectrum (or MS2 sample components). Secondly, the output includes a table of plausible molecular formulas and isotopes, adducts, and oligomers for each component cluster.
[0088] Usage: In one application, the samples may be from experimental studies, such as case-control or longitudinal studies, where statistical tests, such as ANOVA or Welch's t-test, can be applied to the candidate peak areas using the available metadata of the samples to identify statistically significant candidates.
[0089] In another application, samples may come from previously unstudied or understudied matrix types, such as dried blood spots, and the goal may be to identify any new features that occur in most or all of the samples.
[0090] In another application, there may be a single unique sample with the goal of identifying and detecting new compounds in the single unique sample.
[0091] In another application, the sample may come from a laboratory process (such as a water treatment blank) and the goal may be to identify new contaminants.
[0092] Method Steps: 1. Read the raw MS file and collect for each MS2 scan a pair consisting of mass-to-charge ratio (m / z) and retention time (RT) corresponding to the precursor of the MS2 scan (i.e., the ion fragmented in MS2). Normalize the retention times to facilitate alignment of corresponding scans in different samples to obtain pairs (RI, m / z), where RI ("retention index") represents the normalized retention time. While retention times can be normalized in a variety of ways, in one embodiment, retention times are normalized by piecewise linear interpolation between known compounds "spiked" into each sample for calibration.
[0093] 2. Discard pairs that match any entry in the existing MS1 library, whether they are molecular ions, adducts, isotopes, or other derivatives.
[0094] 3. Cluster the remaining pairs into compact features in (RI, m / z) space by any of a number of clustering algorithms known to those skilled in the art, including but not limited to k-means, hierarchical clustering, agglomerative clustering, DBSCAN, Gaussian mixture, or Birch clustering. In one embodiment, the clustering can be performed by a partitional clustering algorithm, as in the following example: a. Put all pairs into a single large cluster. b. Sort all pairs by m / z and split into groups if there are large gaps in m / z. c. Sort each group by RI and split into smaller groups if there are large gaps in RI. d. Sort each group by m / z and split into smaller groups if there are large gaps in m / z. e. Sort each group by RI again and split if there are large gaps in RI. f. Discard groups that are too small, span too much m / z space (e.g., >5 ppm), or span too much RI space (e.g., 250 units). The group at the end of this process is a "feature."
[0095] 4. Obtain all integrated peaks from the (MS1) database of integrated peaks previously derived from samples that are within the range of (RI, m / z) space occupied by the feature.
[0096] 5. For each feature and sample, select the "best" peak from among the peaks obtained. The best peak is the one that is closest to the aggregation RI based on the set of all peaks.
[0097] 6. Discard features with too few samples that have a peak (e.g., require that 5% of samples have a peak).
[0098] 7. Using the pre-calculated tables described below, discard features whose m / z do not have a plausible molecular formula.
[0099] 8. Optionally, calculate the statistical significance of the peak area of each feature with respect to the available metadata. This step uses, for example, a custom code module for each study. Discard non-significant features.
[0100] 9. Re-read the raw MS1 and MS2 files and collect the necessary spectra to complete the analysis of each feature: a. Initialize empty collections of the "best" MS1 and MS2 spectra for each feature. The size of these collections can be limited to 20-50 as an implementation parameter. b. Read the raw files one by one: i. Collect all spectra of the sample. ii. For each feature, find the spectra under the corresponding peak in the sample. If these spectra are the best seen so far, add them to the collection of features and remove spectra that are no longer needed. The "best" spectra are the ones with the highest intensity of the feature ion. This incremental strategy is important for memory management (e.g., not all spectra are stored / kept until all samples have been read).
[0101] 10. Run MS1 analysis for each feature: a. Convert each ion in each MS1 spectrum into a (delta,ratio) pair. Delta is the difference between the m / z of the ion and the m / z of the precursor ion in the same spectrum. m / z=mass to charge ratio, and the ratio is the ratio of the intensity of the ion to the intensity of the precursor ion in the same spectrum. b. The masses of the precursors are used to build a large table of possible adducts, isotopes, and deltas of oligomeric relationships. The computer code that builds this table is constructed automatically using the symbolic algebra outlined below. c. Assign each (delta, ratio) pair to a nearby delta (e.g., ±1 ppm) in a table of possible relationships. These pairs confirm that the corresponding relationship exists in the spectroscopic data. d. Consider the set of pairs assigned for each possible relationship and determine whether there are a sufficient number of pairs and whether the ratio falls within a narrow range. If so, report the relationship. e. Furthermore, the relationship is isotopic, e.g. 13 C. 37 Cl, 81 Br, or 34 For example, S, the median ratio of intensities is used to predict the number of C, Cl, Br, or S in the signature ion based on the binomial distribution and natural abundance of the isotopes. f. Consider the pairs not assigned in (c) and cluster them by delta. If any cluster contains a large number of pairs with similar ratios, report this as a possible relationship of unknown origin. The output of step 10 is a table of isotope, adduct, oligomer and unexplained relationships between signature ions and coincident ions.
[0102] 11. Perform MS2 analysis of each feature (i.e., construct MS2 spectra): a. Cluster the set of all MS2 ions into subsets of ions with likely equal precise m / z according to the instrument-specific measurement error. In one example, the m / z space from 50u to 1000u is divided into approximately 300,000 bins of width 5ppm, and each ion (m / z, intensity) pair is assigned to a bin according to m / z, with each subset consisting of all ions assigned to the same or adjacent bins. b. Find a "normalizing" ion subset: In one preferred embodiment, the normalizing ions are the most intense ions in each MS2 spectrum, but this is not necessarily the case. In each MS2 spectrum, rank the subsets from least to most intense by the intensity of the ions they contain. Sum the ranks of each subset. The subset with the greatest sum becomes the normalizing group. c. For each MS2 spectrum, all intensities are divided by the intensities of a normalized set of the same MS2 spectrum to obtain relative intensities. d. Aggregate the relative intensities of each subset, and eliminate subsets with too small intensities as noise. e. Find the median m / z and median relative intensity for each subset. Report these as the consensus MS2 spectrum. The output of step 11 is a proposed MS2 spectrum of the candidate compound.
[0103] Finding plausible molecular formulas using pre-calculated tables A valid molecular formula is one in which the sum of the masses of the individual atoms is the same as the mass of the feature. A common approach to this problem is that employed, for example, by the ChemCalc website, which, given a mass query, computes a set of formulas apparently on the fly using a branch-and-bound approach. This approach is efficient enough for a small number of queries, and is flexible in that the search parameters (e.g., the maximum number of C) can be specified for each query. However, the methodology allows the search parameters to be fixed for the speedup required for a larger number of queries. This allows precomputation of a large table (in about 10 minutes, once instead of per run) and very fast lookup of that table to answer a query.
[0104] For the purposes of this disclosure, formulas involving einsteinium are not valid. Nearly all compounds in the current library consist of only C, H, N, O, P, S, Cl, F, I, and Br. Masses >1000u are not considered, so compounds of interest are
number
[0105] In general, formulas having, for example, both I and Cl are not necessarily of interest, and analysis of existing libraries shows that in some cases it is sufficient to consider formulas such as:
number
[0106] Find MS1 relationships using auto-generated codes Aspects of the present disclosure provide the ability to efficiently analyze a theoretically unlimited number of previously identified relationships between the m / z of precursor ions and other MS1 ions. These relationships can include isotope adducts, adducts, multiply charged species, and oligomers, as well as isotopic adducts, isotopic oligomers, different isotopes of the same molecule (e.g., 13 C and 37 The mass relationships may include more complex mass relationships such as 1,2,3,4,5,6,7,8,8,9,10,11,12,13,14,15,16,17,18,19,19,20,21,22,23,24,25,26,27,30,28,31,32,29,33,34,35,36,37,38,39,40,41,42,50,51,64,75,86,97,10,11,25,36,37,42,52,78,12,13,26,37,41,51,64,75,87,14,15,28,39,46,79,16,29,48,17,18,29,52,79,18,29,65,86,10,29,30,30,41,52,79,16,29,42,51,64,75,87,10,29,30,42,52,79,10,30,43,54,79,11,29,44,79,12,13,29,54,79,14,29,65,86,10,29,30,44,79,16,29,44,79,16,30,45,50,50,65,86,10,29,47,79,10,30,48,79,10,49,50,50,65,86,10,29,47,79,10,30,48,79,11,29,50,65,86,10,29,47,79
[0107] It would be tedious and error-prone for a human to perform the algebra required to write computer code to evaluate such a large number of relationships, since one must find a symbolic (rather than numerical) solution of one or an algebraic equation for each relationship. The writing of the functions themselves can be automated using symbolic algebra software systems, which accept equations and other mathematical expressions coded as strings and solve the expressions symbolically. For example, "solve 2z=x+y, where y=x+2 for z", the result is a symbolic rather than numerical result "z=x+1". In what follows, m represents the mass of the (uncharged / neutral) precursor biochemical that gives rise to (a portion of) the ions in the hypothetical MS1 scan, specifically, negative ionization is used. Note that m is not the mass of the precursor ion. In fact, this is exactly the point, one objective is to know whether the precursor ion is really the "principal ion" (biochemical-proton) or an adduct or other derivative of the primary ion. Some examples of the use of symbolic algebra are:
[0108] The precursor has a main ion (mH) and another ion 13 C isotope. Algebraically speaking, p=mH represents the mass of the precursor. (In this discussion, H is always the mass of a proton, not a hydrogen atom. In the library, the isotope ion is represented as (mH[C13-1]) and x=m-H+ 13 C- 12 C stands for another ion. In this example:
number
number
[0109] A slightly more difficult case is when the precursor is a dimer (2m-H) and the other ion is a molecular ion (mH). Algebraically,
number
number
[0110] In this case the result is a symbolic result, so the function creates a table with rows having the symbolic expression of delta. When the function is called during the analysis of a particular feature, values are assigned to the variables in the symbolic expression of delta, making delta a numeric value.
[0111] The function builds a table for the analysis based on the mass of the precursor, not the mass of the molecule (which is unknown), so the delta is expressed in terms of p. In this case, m=(p+H) / 2, and the table needs the following rows:
number
[0112] Finally, a more complex example: the precursor is doubly charged. 13 C isotope (m-2H[C13-1]) / 2, other ions are represented as mH-CO2 2 In algebraic form we have a loss of:
number
number
[0113] Now we are left with the problem of converting the molecular species notation into ordinary algebraic expressions. At software build time (i.e., only once), we must decide which relationships (pairs of molecular species) are of interest in the MS1 data analysis. Invalid relationships slow down processing time and can result in false positives, especially when the precursor mass is larger. We then run a program once that itself writes computer code to efficiently evaluate many hypothetical relationships between the precursor and other ions in the MS1 data.
[0114] Applying the Output One goal is to generate library entries consisting of m / z, RI, MS1 spectra, and MS2 spectra, thus identifying plausible molecular formulas, adducts, isotopes, and oligomers.
[0115] Having a reasonable molecular formula is important for several reasons. If none is present, this suggests that the feature is not a molecular (m+H or mH) ion, perhaps an isotope or oligomer, or perhaps simply noise. Also, if MS1 data analysis shows a feature repeatedly as something other than a molecular ion, this also counts against the creation of a library entry. Finally, because metabolites are C-rich, reasonably large metabolites are unlikely to be C-rich unless the precursor ion intensity is very low. 13 It should show C ions.
[0116] example: In one example, this methodology was used to analyze 59 mouse plasma samples, obtained from three groups of mice corresponding to three genotypes at the HAL (histidine ammonia lyase) locus: reference (REF), heterozygous (HET), or knockout (VAR).
[0117] To assess the validity of the proposed compounds, 15 compounds were subjected to detailed manual review by human experts, which showed that 11 of the 15 compounds (73.3%) were likely to be true novel compounds.
[0118] Accordingly, aspects of the present disclosure provide methods of analyzing metabolomic data from the LC / tandem MS system disclosed herein. In addition to providing suitable apparatus and methods, aspects of the present disclosure also provide associated computer program products for performing the functions / operations / steps disclosed herein, for example in the form of a non-transitory computer readable storage medium (i.e., memory device 140 of FIG. 1) having certain computer readable program code portions stored thereon, which upon execution by the processor device 130 causes the apparatus to perform at least the steps disclosed herein. In this regard, it will be understood that each block or step of the methodology, or combinations of blocks / steps in the methodology, can be implemented by suitable computer program instructions executed by the processor device 130. These computer program instructions can be loaded into a computer device or other programmable apparatus for performing the functions specified in the methodology or related to the methods disclosed herein. These computer program instructions can also be stored in a computer readable memory (i.e., memory device 140) such that they are accessible in a specific manner by a computing device or other programmable apparatus, such that the executable instructions stored in the computer readable memory can cause or facilitate the operation of an article of manufacture capable of directing or executing instructions that implement functions associated with the methods specified in the methodologies or otherwise disclosed herein. The computer program instructions can also be loaded into a computing device or other programmable apparatus and caused to execute a series of operational steps on the computing device or other programmable apparatus to produce a computer implemented process such that the instructions executed by the computing device or other programmable apparatus provide or otherwise direct the appropriate steps for implementing the functions / steps associated with the methods specified in the methodologies or otherwise disclosed herein.It will also be understood that each step of the methodology, or combinations of steps in the methodology, can be implemented by a dedicated hardware-based computer system that performs the specified functions or steps, or a combination of dedicated hardware and computer instructions (software).
[0119] Many modifications and other embodiments of the inventions described herein will occur to those skilled in the art having the benefit of the teachings presented in the above description and the associated drawings to which these disclosed embodiments pertain. Thus, it is understood that the embodiments of the invention should not be limited to the specific embodiments disclosed, and that modifications and other embodiments are intended to be included within the scope of the invention. Furthermore, while the above description and the associated drawings describe exemplary embodiments in connection with certain exemplary combinations of elements and / or functions, it is to be understood that different combinations of elements and / or functions may be provided by alternative embodiments without departing from the scope of the disclosure. In this regard, for example, combinations of elements and / or functions different from those explicitly described above are also contemplated within the scope of the disclosure. Although specific terms have been employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.
[0120] Terms such as first, second, etc. may be used herein to describe various steps or calculations, but it should be understood that these steps or calculations should not be limited by these terms. These terms are used only to distinguish one operation or calculation from another operation or calculation. For example, a first calculation can be referred to as a second calculation, and similarly, a second step can be referred to as a first step, without departing from the scope of the present disclosure. The term "and / or" and the " / " symbol as used herein include any and all combinations of one or more of the associated listed items.
[0121] As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly dictates otherwise. It will be further understood that the terms "comprising," "including," and "including" as used herein specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Accordingly, the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.
Claims
1. A method of analyzing data for one or more samples, the data for each sample being obtained from a component separation and tandem mass spectrometry system comprising a component separator portion comprising a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, and a mass spectrometry portion comprising a first mass spectrometry functional portion (MS1) and a second mass spectrometry functional portion (MS2), the data from MS1 comprising MS1 sample components and the data from MS2 comprising MS2 sample components; The method further comprising: analyzing a plurality of MS2 sample component data sets for each sample to determine a mass to charge ratio (m / z) and a retention index (RI) of a precursor ion of each MS2 sample component; comparing the m / z and RI of the precursor ion of each MS2 sample component with the mass-to-charge ratio and retention index of precursor ions of known compounds, removing any MS2 sample component that corresponds to one of the known compounds from the data set, and defining the MS2 sample components remaining in the data set as candidate MS2 sample components; forming component clusters over the candidate MS2 sample components, each component cluster including candidate MS2 sample components having precursor ion m / z and RI within a respective range of each component cluster; For each sample having candidate M2 sample components in each component cluster, acquiring one or more MS1 sample components within the m / z and RI range of a respective precursor ion of the component cluster; for each sample having one or more MS1 sample components in each component cluster, determining at most one consensus MS1 sample component representing each sample in each component cluster by aggregating one or more MS1 sample components within a range of m / z and RI of a respective precursor ion, and associating the consensus MS1 sample component with a corresponding consensus MS2 sample component, consensus precursor ion m / z, and consensus RI determined by aggregating MS2 sample components and the m / z and RI of associated precursor ions in the component cluster; determining, for each component cluster, whether the consensus MS1 sample components exhibit an ionization product or a derivative relationship to an ionization product; grouping the constituent clusters according to a consensus RI and selecting, from each group of constituent clusters, one or more constituent clusters that are candidate constituent clusters; correlating the consensus precursor ion m / z, the consensus RI, the consensus MS1 sample component, and the consensus MS2 sample component of the candidate component cluster with unnamed compounds in the sample; A method comprising:
2. 2. The method of claim 1, wherein removing any MS2 sample components corresponding to the known compounds from the dataset comprises removing from the dataset any of known ionization products, background artifacts, chromatographic artifacts, derivative relationships of known ionization products, derivative relationships of mass relationships of known ionization products, derivative relationships of background artifacts, or derivative relationships of chromatographic artifacts corresponding to one of the known compounds.
3. 3. The method of claim 2, wherein removing the derivative relationships of the known ionization products or the derivative relationships of the mass relationships of the known ionization products from the dataset comprises removing any of in-source fragments, adducts, isotopes, dimers, oligomers, multiply charged species, isotopic adducts, or isotopic oligomers from the dataset.
4. 10. The method of claim 1, comprising determining the mass to charge ratio and retention index of a precursor ion of a known compound from an MS1 sample component of the known compound.
5. 5. The method of claim 4, further comprising associating any of the m / z, RI, MS1 sample component, and MS2 sample component of the precursor ion of each known compound with the known compound in a database comprising an ion data repository.
6. 2. The method of claim 1, comprising associating the m / z, RI, MS1 sample component, and MS2 sample component of the precursor ion of each candidate component cluster with an unnamed compound in a database comprising an ion data repository.
7. 2. The method of claim 1, comprising determining the retention index by normalizing the retention time of each MS2 sample component.
8. The method of claim 1 , wherein forming component clusters across the candidate MS2 sample components comprises forming component clusters across the candidate MS2 sample components using divisive clustering or agglomerative clustering.
9. 2. The method of claim 1, further comprising sorting the candidate MS2 sample components by precursor ion mass prior to forming component clusters across the candidate MS2 sample components.
10. The method of claim 9 , comprising sorting the component clusters by retention index to generate secondary component clusters.
11. The method of claim 1 , wherein the step of forming component clusters across the candidate MS2 sample components is repeated two or more times.
12. The method of claim 1 , wherein comparing the consensus MS1 sample components of each component cluster across the component clusters is performed across multiple samples.
13. 13. The method of claim 12, comprising removing any component cluster that is present in less than 5% of the plurality of samples.
14. 2. The method of claim 1, further comprising analyzing a plurality of possible derivative relationships or possible mass relationships of the MS1 sample components prior to associating the consensus MS1 sample components with the m / z of the corresponding consensus precursor ion.
15. 15. The method of claim 14, wherein analyzing the plurality of possible derivative relationships or the plurality of possible mass relationships comprises creating computer code for analyzing the plurality of possible derivative relationships or possible mass relationships in one run using symbolic algebra to analyze a specified set of derivative relationships or mass relationships.
16. The method of claim 1 , further comprising eliminating component clusters that do not have a valid molecular formula prior to grouping the component clusters.
17. 10. The method of claim 1, further comprising removing component clusters without statistical significance from metadata of data of the one or more samples being analyzed prior to grouping the component clusters.
18. 10. The method of claim 1, comprising prioritizing candidate component clusters that have statistical significance with respect to metadata of data of one or more samples analyzed as unnamed candidate compounds.
19. 1. An apparatus for analyzing data of one or more samples, comprising: Data for each sample is obtained from a component separation and tandem mass spectrometry system comprising a component separator section including a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, and a mass spectrometry section including a first mass spectrometry functional section (MS1) and a second mass spectrometry functional section (MS2); a processor; and a memory storing executable instructions which, when executed by the processor, cause an apparatus to perform at least the method steps of any one of claims 1 to 18; An apparatus comprising:
20. 1. A computer program product for analyzing data for one or more samples, comprising: Data for each sample is obtained from a component separation and tandem mass spectrometry system comprising a component separator section including a liquid chromatograph, a gas chromatograph, a supercritical fluid chromatograph, or a capillary electrophoresis analyzer, and a mass spectrometry section including a first mass spectrometry functional section (MS1) and a second mass spectrometry functional section (MS2); 19. A computer program product, the computer program product comprising at least one non-transitory computer readable storage medium having computer readable program code stored thereon, the computer readable program code comprising program code for performing the steps of the method according to any one of claims 1 to 18.
Citation Information
Patent Citations
Mass analyzing system
JP2008170260A
How to identify fungi
JP2016537964A
Method for analyzing small molecule components of complex mixtures in multiple sample processes, as well as related devices and computer program products
JP2017517731A
Techniques for sample analysis using consensus libraries
WO2019229723A1
