Identification and use of biological markers for diagnosis and treatment monitoring

Abstract

This patent relates to a system and method for quantitatively measuring the glycomic index, genomic index, proteomic index, metabolomic index and / or lipidomic index of a biological sample: obtaining clinical indicators related to the subject from which the one or more biological samples are derived; determining one or more relationships between one or more of the following: (i) one or more quantitative glycomic indexes, genomic indexes, proteomic indexes, metabolomic indexes and lipidomic indexes, (ii) preset ranges related to one or more quantitative glycomic indexes, genomic indexes, proteomic indexes, metabolomic indexes and lipidomic indexes, and (iii) the obtained clinical indicators. It also involves identifying one or more biomarkers based on one or more deterministic relationships that meet preset importance criteria; and / or determining the health classification status of a health classification based on one or more effective biomarkers.

Description

Technical Field

[0001] This patent generally relates to the diagnosis and treatment of health conditions. In some specific embodiments, this patent relates to new systems and methods for associating biological indicators with health classifications, health states, treatment effects, health improvements or deviations, etc. Background Art

[0002] Timely diagnosis and treatment of health conditions are of great importance to the healthcare community. Traditional methods for diagnosing and treating health conditions fall short in terms of accuracy and precision. In particular, traditional methods for interpreting mass spectra obtained from biological samples are susceptible to human error. Human input often introduces biases that can affect the conclusions of mass spectrometry interpretation. There is a need for new systems and methods to improve the reliability, accuracy, and precision of mass spectrometry interpretation by making unbiased and continuously validated decisions in an intelligent environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0003] Figure 1A is a schematic diagram of an example system, which, according to one or more embodiments of the present patent, is capable of identifying one or more biological indicators associated with one or more health classifications and predictively diagnosing one or more health states of one or more subjects based on the biological indicators.

[0004] Figure 1B is a schematic diagram of an example system capable of quantifying biological indicators using a peak integration platform, identifying one or more biological indicators associated with one or more health categories, and predictively diagnosing one or more health states of one or more subjects based on the biological indicators.

[0005] Figure 1C Schematic diagram of mass spectrometry that can analyze biological samples.

[0006] Figure 1D is a schematic diagram of a peak waveform generated based on the mass spectrometry of an analyzable biological sample.

[0007] Figure IE shows the integral of the peak waveform shown in Figure ID.

[0008] Figure 2 is a flow chart of an example method for determining one or more biological indicators as one or more biomarkers.

[0009] Figure 3 is a schematic diagram of an example system capable of performing one or more automated unbiased deep learning operations to determine biomarkers.

[0010] Figure 4 is a flow chart of an example method for performing automated unbiased deep learning operations to determine biomarkers.

[0011] Figure 5 is a diagram of an example system for diagnosing a disease in a subject based on biomarkers.

[0012] Figure 6 is a graphical representation of exemplary changes in the ratio of immunoglobulin G (IgG) glycopeptides in plasma samples from breast cancer patients relative to plasma samples from a control group.

[0013] Figure 7 Shown are changes in IgG glycopeptide ratios in plasma samples from patients with primary sclerosing cholangitis (PSC) and primary biliary cirrhosis (PBC) relative to plasma samples from healthy donors.

[0014] Figures 8A-8C Individual discriminant analysis data for IgG, IgA, and IgM glycopeptides in plasma samples from PSC and PBC samples relative to plasma samples from healthy donors are shown, respectively.

[0015] Fig. 9 Shown are combined discriminant analysis data for IgG, IgA, and IgM glycopeptides in plasma samples from PSC and PBC patients versus plasma samples from healthy donors.

[0016] Detailed description

[0017] definition

[0018] In this specification, the following words and phrases generally have the following meanings unless the context dictates otherwise.

[0019] The term "biological sample" refers to any biological fluid, cell, tissue, organ, or any part of one or more of the foregoing, or any combination of any one or more of the foregoing. For example, a "biological sample" may include one or more tissue sections obtained by biopsy; cells undergoing or suitable for tissue culture; samples of saliva, tears, sputum, sweat, mucus, feces, gastric juice, peritoneal fluid, amniotic fluid, cyst fluid, peritoneal fluid, spinal fluid, urine, synovial fluid, whole blood, serum, plasma, pancreatic juice, breast milk, lung lavage fluid, bone marrow, gastric acid, bile, synovial fluid, semen, pus, aqueous humor, exudate, etc.; or any other biological substance, or any part of any one or more of the foregoing, or any combination of any one or more of the foregoing.

[0020] The term "biomarker" refers to a unique biological or biologically derived indicator of one or more processes, events, conditions, or any combination thereof. In general, biological indicators and indicators derived from biology are detectable, quantifiable, and / or measurable. For example, a biomarker may comprise one or more measurable molecules or substances that are produced by, associated with, or obtained from a subject, and whose presence indicates another quality, such as one or more processes, events, conditions, or any combination thereof. A biomarker may include any one or more biological molecules (alone or in combination), or any one or more fragments of biological molecules (alone or in combination), the detected presence, amount (absolute amount, proportional amount, relative amount, or other form), measure, or one or more such changes in presence, amount, or measure, associated with one or more specific health conditions. For example, biomarkers include, but are not limited to, one or more biological molecules: nucleotides, amino acids, fatty acids, steroids, antibodies, hormones, peptides, proteins, carbohydrates, and the like. Other examples include one or more glycosylated peptide fragments, lipoproteins, and the like. A biomarker may indicate a health condition, such as the presence, onset, stage, or state of, or risk for, one or more diseases, infections, syndromes, symptoms, or other conditions.

[0021] The term "glycan" refers to the carbohydrate portion of a glycoconjugate such as a glycopeptide, glycoprotein, glycolipid, or proteoglycan.

[0022] The term "glycoform" refers to the unique primary, secondary, tertiary, and quaternary structures of a protein to which specific structural glycans are attached.

[0023] The term "glycosylated peptide fragment" refers to a glycosylated peptide (or glycopeptide) obtained by fragmenting a glycosylated peptide by one or more proteases, and its amino acid sequence is partially identical but not completely identical to the glycosylated peptide.

[0024] The term "Multiple Reaction Monitoring Mass Spectrometry (MRM-MS)" refers to a highly sensitive, highly selective, targeted quantification method for proteins / peptides in biological samples. Unlike traditional mass spectrometry, MRM-MS is highly selective (targeted) and researchers can fine-tune the instrument to find specific target peptides / protein fragments. MRM has higher sensitivity, higher specificity, faster analysis, and a larger limit of quantification, and can be used to detect specific peptides / protein fragments, such as potential biomarkers. MRM-MS includes triple quadrupole (QQQ) mass spectrometers or quadrupole time-of-flight (qTOF) mass spectrometers.

[0025] The term "protease" refers to an enzyme that can proteolyze or decompose proteins into smaller polypeptides or amino acids. Examples of proteases include, but are not limited to, one or more of serine proteases, threonine proteases, cysteine ​​proteases, aspartic proteases, glutamic proteases, metalloproteases, asparagine peptide cleavage enzymes, and any combination thereof.

[0026] The term "subject" refers to a mammal. Non-limiting examples of mammals include humans, non-human primates, mice, rats, dogs, cats, horses or cattle, etc. Mammals other than humans are more advantageous as subjects representing animal models of disease, pre-disease or pre-disease states. The subject can be male or female. The subject can be previously determined to have a certain disease or symptom, or can have or are receiving a therapeutic intervention for a certain disease or symptom. As an alternative, the subject can also not have been previously diagnosed with a disease or symptom. For example, the subject can show one or more risk factors for a disease or symptom, or does not show a disease risk factor, or has no expression of a disease or symptom. The subject can also be suffering from a disease or symptom or be at risk of suffering from a disease or symptom.

[0027] The terms "treatment" or "treating" refer to any treatment of a disease or condition in a subject, such as a mammal, including: (1) preventing the disease or condition, i.e., preventing clinical symptoms from developing; (2) inhibiting the disease or condition, i.e., arresting or inhibiting the development of clinical symptoms; and / or (3) ameliorating the disease or condition, i.e., causing regression of clinical symptoms.

[0028] The singular nouns “a”, “an”, “the” and “said” used in this specification include plural nouns unless the context clearly indicates otherwise.

[0029] system

[0030] Figure 1A A diagram of an example system is described, which can identify biological indicators associated with health classification and predictively diagnose the health status of a subject based on the biological indicators. As shown in the figure, the system 100 includes a computer readable medium 102, a glycomic indicator quantification system 104, a genomic indicator quantification system 106, a proteomics indicator quantification system 108, a metabolomics indicator quantification system 110, a lipidomics indicator quantification system 112, a clinical indicator generation system 114, an automatic unbiased machine learning diagnosis system 116 and a diagnosis result delivery system 118.

[0031] Computer readable medium 102 refers to various applicable technologies. For example, computer readable medium 102 can be used to form a network or a part of a network. When two components are located on a device at the same time, computer readable medium 102 may include a bus or other data channel or plane. In the case where a first component is located on one device and a second component is located on a different device, computer readable medium 102 may include a wireless or wired background network or LAN. If applicable, computer readable medium 102 may also include relevant parts of a WAN or other network.

[0032] As used in this specification, "computer-readable media" includes all media that are statutory (e.g., pursuant to 35 USC 101); specifically, all media that are not statutory in nature are excluded if necessary for the validity of a claim that includes computer-readable media. Known statutory computer-readable media include hardware (e.g., registers, random access memory (RAM), non-volatile (NV) storage devices, to name a few), and may or may not be limited to hardware.

[0033] The computer readable medium 102 described in this specification or a portion thereof, as well as other systems, interfaces, engines, data storage repositories and other devices, can be implemented as a computer system, multiple computer systems or a part of a computer system or multiple computer systems. Typically, a computer system includes a processor, a memory, a non-volatile memory and an interface. A typical computer system typically includes at least a processor, a memory and a device (such as a bus) connecting the memory to the processor. The processor can be a general-purpose central processing unit (CPU) such as a microprocessor, or a special-purpose processor such as a microcontroller.

[0034] Non-limiting examples of memory may include random access memory (RAM) such as dynamic RAM (DRAM) and static RAM (SRAM). The memory may be local, remote or distributed. The bus may also connect the processor to non-volatile memory. Non-volatile memory is typically a floppy or hard disk, a magneto-optical disk, an optical disk, a read-only memory (ROM) such as a CD-ROM, EPROM or EEPROM, a magnetic or optical card, or other form of memory for large amounts of data. Some of this data is typically written to the memory by a direct memory access program during the execution of software by the computer system. Non-volatile memory may be local, remote or distributed. Non-volatile memory is optional because the system can be created using all available data in the memory.

[0035] Software is typically stored in non-volatile memory. In fact, for large programs, it may not be possible to store the entire program in memory. However, it should be understood that to run the software, it can be moved to a computer-readable location suitable for processing if necessary; for the purposes of illustration, the location is referred to as memory in this specification. Even if the software is moved to memory for execution, the processor will typically use hardware registers to store values ​​related to the software, and ideally, local caches can be used to speed up execution. In this specification, when a software program is referred to as "applied in a computer-readable storage medium", it is assumed that the software program is stored in an available known or convenient location (from non-volatile memory to hardware registers). When at least one value associated with the program is stored in a register that can be read by the processor, the processor is considered to be "capable of executing the program."

[0036] In one example of operation, a computer system may be controlled by operating system software, which is a file management system software program including, for example, a disk operating system. An example of operating system software with associated file management system software is Microsoft Windows from Microsoft Corporation of Redmond, Washington. A family of operating systems and their associated file management systems. Another example of operating system software with associated file management system software is the Linux operating system and its associated file management system. The file management system causes the processor to perform various operations required by the operating system to input and output data and store data in memory, including storing files on non-volatile memory.

[0037] The bus can also connect the processor to the interface. The interface can include one or more input and / or output (I / O) devices. Non-limiting examples of I / O devices can include keyboards, mice or other pointing devices, disk drives, printers, scanners and other I / O devices, as well as display devices. Non-limiting examples of display devices can include cathode ray tubes (CRTs), liquid crystal displays (LCDs) or other available known or convenient display devices. The interface can include one or more modems or network interfaces. It should be understood that a modem or network interface can be considered as part of a computer system. The interface can include an analog modem, an ISDN modem, a cable modem, a token ring interface, a satellite transmission interface (e.g., "satellite live Internet") or other interfaces for connecting a computer system to other computer systems. The interface enables a computer system to be connected to other devices in a network.

[0038] The computer system may be compatible with, or be part of, or be implemented by a cloud-based computing system. As used in this specification, a cloud-based computing system is a system that provides virtualized computing resources, software, and / or information to client devices. Computing resources, software, and / or information are virtualized by maintaining centralized services and resources, which can be accessed by edge devices through communication interfaces such as networks. "Cloud" may be a marketing term, and for illustrative purposes, it may include any of the networks described in this specification. A cloud-based computing system may involve service subscriptions or use a utility pricing model. A user may access a cloud-based computing system protocol through a web browser or other container application located on their client device.

[0039] The computer system may be implemented as an engine, a portion of an engine, or may be implemented via multiple engines. The engine referred to in this specification includes at least two components: (1) a dedicated or shared processor; and (2) hardware, firmware, and / or software modules executed by the processor. Based on a particular implementation or other considerations, the engine may be centralized or its functions may be decentralized. The engine may include dedicated hardware, firmware, or software in a computer-readable medium for execution by the processor. As shown, the processor uses implemented data structures and methods to convert data into new data.

[0040] The engine described herein or the engine by which the systems and devices described herein can be implemented can be a cloud-based engine. The cloud-based engine described herein is an engine that can use a cloud-based computing system to run applications and / or functions. All or part of these applications and / or functions can be distributed among multiple computing devices and are not limited to a single computing device. In some embodiments, a cloud-based engine can execute functions and / or modules that an end user accesses via a web browser or container application without having to install these functions and / or modules locally on the end user's computing device.

[0041] The data repository described herein includes a repository with any available data organization, wherein the data organization includes a table, a comma separated value (CSV) file, a traditional database (e.g., SQL), or other applicable known or conveniently managed format. For example, the data repository can be implemented as software, which is embodied in a physical computer-readable medium on a general-purpose or special-purpose machine, in firmware, in hardware, in a combination thereof, or in an available known or convenient device or system. Although the physical location and other characteristics of data repository-related components are not important for understanding the technology described herein, data repository-related components such as database interfaces can be considered "part" of the data repository, part of other system components, or a combination thereof.

[0042] A data repository may include a data structure. The data structures described herein are related to a specific way of storing and organizing data in a computer, which allows the data structure to be used efficiently in specific circumstances. Data structures are generally based on the ability of a computer to extract and store data at a specified address (i.e., a bit string that can itself be stored in the memory and manipulated by a program) anywhere in its memory. Therefore, some data structures are based on calculating the addresses of data items using arithmetic operations; while other data structures are based on storing the addresses of data items in the structure itself. Many data structures use these two principles, sometimes in combination in a meaningful way. The implementation of a data structure usually requires writing a set of programs to create and manipulate instances of the structure. The data repository described herein may be a cloud-based data repository. A cloud-based data repository is a data repository compatible with a cloud-based computing system and engine.

[0043] Reference again Figure 1A In an example, a glycomic index quantification system 104 is connected to the computer-readable medium 102. The glycomic index quantification system 104 represents a system that can quantify the glycomic index of a biological sample, and the system can also provide information on the glycomic index quantification results to the computer-readable medium 102. In order to quantify the glycomic index in the biological sample, the glycomic index quantification system 104 can be controlled by the entity (such as a hospital) that collects the biological sample, or it can not be controlled by it. Glycomic indicators can include the number and changes in the number of glycosylated proteins in the biological sample, the number and changes in the number of types of glycosylated peptide fragments fragmented by glycosylated proteins, and the source of the biological sample. In one embodiment, the glycomic index quantification system 104 runs continuously, so that whenever a new biological sample is obtained, its quantitative results can be obtained.

[0044] In some embodiments, the biological sample is from one or more previous studies, spanning 1 to 50 years or even longer. In some embodiments, the study also includes various other clinical indicators and previously known information, such as the subject's age, height, weight, race, medical history, etc. This additional information can be used to associate the subject with a disease or symptom. In some embodiments, the biological sample is one or more clinical samples collected in advance from the subject.

[0045] In one embodiment, the biological sample isolated from the subject is body tissue, saliva, tears, sputum, spinal fluid, urine, synovial fluid, whole blood, serum or plasma. In another embodiment, the biological sample isolated from the subject is whole blood, serum or plasma. In other embodiments, the subject is a mammal. In other embodiments, the subject is a human.

[0046] In one embodiment, the glycosylated proteins used to quantify glycomic indicators are one or more of the following: α-1-acid glycoprotein, α-1-antitrypsin, α-1B-glycoprotein, α-2-HS-glycoprotein, α-2-macroglobulin antithrombin-III, apolipoprotein B-100, apolipoprotein D, apolipoprotein F, β-2-glycoprotein-1, ceruloplasmin, fetuin, fibrinogen, immunoglobulin (Ig) A, IgG, IgM, haptoglobin, heme, histidine-rich glycoprotein, kininogen-1, serum transferrin, transferrin, vitronectin and zinc-α-2-glycoprotein.

[0047] In one embodiment, the glycosylated peptide fragments used to quantify glycomic indicators are one or more of O-glycosylation and N-glycosylation. In another embodiment, the glycosylated peptide fragments used to quantify glycomic indicators have an average length of 5 to 50 amino acid residues. In another embodiment, the average length of the glycosylated peptide fragments is about 5 to about 45, or about 5 to about 40, or about 5 to about 35, or about 5 to about 30, or about 5 to about 25, or about 5 to about 20, or about 5 to about 15, or about 5 to about 10, or about 10 to about 50, or about 10 to about 45, or about 10 to about 40, or about 10 to about 35, or about 10 to about 30, or about 10 to about 25, or about 10 to about 20, or about 10 to about 15, or about 15 to about 45, or about 15 to about 40, or about 15 to about 35, or about 15 to about 30, or about 15 to about 25, or about 15 to about 20 amino acid residues. In one embodiment, the average length of the glycosylated peptide fragments is about 15 amino acid residues. In another embodiment, the average length of the glycosylated peptide fragments is about 10 amino acid residues. In another embodiment, the average length of the glycosylated peptide fragments is about 5 amino acid residues.

[0048] In one embodiment, one or more proteases are used to fragment the glycosylated protein. In one embodiment, the one or more proteases include serine proteases, threonine proteases, cysteine ​​proteases, aspartic proteases, glutamic proteases, metalloproteases, asparagine peptide lyases or combinations thereof. Some representative examples of proteases include, but are not limited to, trypsin, chymotrypsin, endoproteases, Asp-N, Arg-C, Glu-C, Lys-C, pepsin, thermophilic bacteria, esterases, papain, proteinase K, subtilisin, clostripain, carboxypeptidase, etc. In another embodiment, in the method as described above provided by this patent, the one or more proteases include at least two proteases. In another embodiment, the fragmentation and quantification of glycosylated proteins use liquid chromatography-mass spectrometry (LC-MS) technology combined with multiple reaction monitoring mass spectrometry (MRM-MS), and hundreds of glycosylated fragments (and their parent proteins) can be quantified in a single LC / MRM-MS analysis. This patented advanced mass spectrometry technology provides an efficient ion source, higher resolution, faster separations, and detectors with higher dynamic range, enabling a wide range of non-targeted measurements while retaining the benefits of targeted measurements.

[0049] The mass spectrometry described in this patent can analyze multiple glycosylated proteins at a time. For example, using a mass spectrometer, at least more than 50, or at least more than 60, or at least more than 70, or at least more than 80, or at least more than 90, or at least more than 100, or at least more than 110 or more than 120 glycosylated proteins can be analyzed at a time.

[0050] In one embodiment, the mass spectrometry of the present invention uses a QQQ or qTOF mass spectrometer. In another embodiment, the mass spectrometry of the present invention provides data with a mass accuracy of 10 ppm or more, or 5 ppm or more, or 2 ppm or more, or 1 ppm or more, or 0.5 ppm or more, or 0.2 ppm or more, or 0.1 ppm or more, and a resolution of 5,000 or more, or 10,000 or more, or 25,000 or more, or 50,000 or more, or 100,000 or more.

[0051] exist Figure 1AIn the example of , the genomic index quantification system 106 is connected to the computer readable medium 102. The genomic index quantification system 106 represents a system for quantifying the genomic index of the biological sample, and the system can also provide information of the genomic index quantification result to the computer readable medium 102. In order to quantify the genomic index of the biological sample, the genomic index quantification system 106 can be controlled by the entity (e.g., hospital) that collects the biological sample, or it can not be controlled by it. In implementation, the genomic index can include the genome sequence of DNA or RNA extracted from the biological sample. The method of DNA (RNA) sequencing is not particularly limited. In implementation, the method can include Maxam-Gilbert sequencing, chain termination method, massively parallel signature sequencing (MPSS), polymerase clone sequencing, 454 pyrophosphate sequencing, illumination sequencing, SOLid sequencing, ion torrent semiconductor sequencing, DNA nanoball sequencing, helicopter microscope single molecule sequencing, single molecule real-time (SMRT) sequencing, nanopore DNA sequencing, tunnel current DNA sequencing, hybridization sequencing, mass spectrometry sequencing, microfluidic Sanger sequencing, RNAP sequencing and in vitro virus high-throughput sequencing. In implementation, similar to the glycomics index quantification system 104 , the genomics index quantification system 106 is continuously run to update data.

[0052] exist Figure 1A In the example of , the proteomics index quantification system 108 is connected to the computer-readable medium 102. The proteomics index quantification system 108 represents a system for quantifying the proteomics index of a biological sample, and the system can also provide the computer-readable medium 102 with relevant information of the quantitative results of the proteomics index. In order to quantify the proteomics index from the biological sample, the proteomics index quantification system 108 can be controlled by the entity (e.g., a hospital) that collects the biological sample, or it can not be controlled by it. In implementation, the proteomics index can include the number and quantity changes of various proteins in the biological sample, as well as the source of the biological sample. The method for detecting and / or quantifying proteins is not particularly limited. In implementation, the method can include enzyme-linked immunosorbent assay (ELISA), Western blotting, Edman degradation, matrix-assisted laser desorption / ionization (MALDI), electrospray ionization (ESI), mass spectrometry immunoassay (MSIA) and anti-peptide antibody capture stable isotope standard method (SISCAPA). In implementation, similar to the glycomics index quantification system 104, the proteomics index quantification system 108 runs continuously to update the data.

[0053] exist Figure 1AIn the example of, the metabolic index quantitative system 110 is connected to the computer readable medium 102. The metabolic index quantitative system 110 represents a system for quantifying the metabolic index of a biological sample, and the system can also provide the computer readable medium 102 with relevant information of the metabolic index quantitative results. In order to quantify the metabolic index from the biological sample, the metabolic index quantitative system 110 can be controlled by the entity (such as a hospital) that collects the biological sample, or it can not be controlled by it. In implementation, the metabolic index may include the number and quantity changes of the metabolites and / or byproducts (including sugars, nucleotides and amino acids) of the subject, the biological state of the subject caused by metabolism, and the source of the biological sample, etc. The metabolic index can be quantified by any known method, such as liquid chromatography-mass spectrometry (LC-MS) technology using multiple reaction monitoring mass spectrometry (MRM-MS), etc. In implementation, similar to the glycomics index quantitative system 104, the metabolic index directional system 110 is operated in a continuous manner to update the data.

[0054] exist Figure 1A In the example of , the lipidomics index quantification system 112 is connected to the computer-readable medium 102. The lipidomics index quantification system 112 represents a system for quantifying lipidomics indexes of biological samples, which can also provide relevant information of the quantitative results of lipidomics to the computer-readable medium 102. In order to quantify lipidomics indexes from biological samples, the lipidomics index quantification system 112 may be controlled by the entity (e.g., a hospital) that collects the biological samples, or may not be controlled by it. In implementation, the lipidomics index may include the number and quantity changes of any lipids, such as acyglycerol, wax, ceramide, phospholipid, sphingomyelin, glycerophospholipid, sphingolipid, glyceroglycolipid, lipoprotein, sulfatide, fatty acid, terpenoid, steroid and carotenoid, as well as the source of the biological sample from which the lipids are obtained. In implementation, similar to the glycomic index quantification system 104, the lipidomics index quantification system 112 runs in a continuous manner to update data.

[0055] exist Figure 1AIn the example of , the clinical indicator generation system 114 is connected to the computer-readable medium 102. The clinical indicator generation system 114 represents a system that generates clinical indicators for biological samples and provides relevant information of the clinical indicators to the computer-readable medium 102. In order to generate clinical indicators for subjects, the clinical indicator generation system 114 may be controlled by an entity (e.g., a hospital) that collects clinical data of the subject, or may not be controlled by it. In implementation, the clinical indicators may include any quantifiable and / or non-quantifiable data obtained by examining the subject (e.g., heart rate, blood pressure, blood type, body temperature, skin color, eye color, blood glucose concentration, weight, height, current perceived health classification status, etc.), and any data obtained by asking the subject or from the medical record (e.g., lifestyle habits, including food, sleep and wake-up time, amount of exercise and frequency of exercise, amount of smoking and frequency of smoking, alcohol consumption and frequency of drinking, allergies, medications taken, medical history, race, pain and causes of pain, etc.). In implementation, similar to the glycomics indicator quantification system 104, the clinical indicator generation system 114 runs in a continuous manner to update the data.

[0056] Although specific implementations are included in clinical and laboratory ecosystems, it should be understood that other indicator generation systems may also be used, such as social media indicator generation systems that obtain subject data from social media, behavioral indicator generation systems that obtain online activity data from various sources, and government record indicator generation systems that obtain publicly available data from government websites. The larger the data sample size, the more different data can be included in the health classification indicator.

[0057] exist Figure 1A In the example of , the automatic unbiased machine learning diagnostic system 116 is connected to the computer readable medium 102. The automatic unbiased machine learning diagnostic system 116 represents a system controlled by an entity (e.g., a hospital) that is responsible for identifying one or more biological indicators associated with a specific health classification. The entity may be the same entity as the entity controlling the glycomics indicator quantification system 104, the genomics indicator quantification system 106, the proteomics indicator quantification system 108, the metabolomics indicator quantification system 110, the lipidomics indicator quantification system 112, and the clinical indicator generation system 114, or may not be the same entity.

[0058] In one embodiment, the automatic unbiased machine learning diagnostic system 116 is capable of automatically determining the abundance or absence of one or more quantifiable biological indicators as biomarkers associated with a specific health classification, and / or automatically determining the presence or absence of one or more non-quantifiable biological indicators as biomarkers associated with a specific health classification. Depending on the specific embodiment or other considerations, the biological indicator determined as a biomarker can be a scalar value or range of values ​​of a biological indicator, or a combination of two or more biological indicators (e.g., a ratio of two biological indicators and a vector of two or more biological indicators). For example, a range of metabolites indicating a health condition (e.g., above a threshold or between a lower threshold and a higher threshold). In another example, a specific ratio or range of ratios of a certain type of glycopeptide to a certain type of lipid can indicate a health condition. In another example, a combination of some quantifiable biological indicators and positive non-quantifiable indicators (e.g., non-smokers) that exceed a certain threshold can be used as a biomarker.

[0059] In one specific embodiment, the automatic unbiased machine learning diagnostic system 116 prohibits or restricts users from making changes to the indicator settings used for a specific data calculation process to ensure that the machine's automatic calculation is free of human intervention (e.g., no human bias). This is because, to an observer (e.g., a researcher), some biomarkers appear to be irrelevant, and this human bias can make it more difficult to find biomarkers for health classifications. For example, in the automatic unbiased machine learning diagnostic system 116, at least in the initial stages of the calculation, each biological indicator included in the automatic unbiased machine learning diagnostic system 116 has an equal weight. In other words, in the initial stages of the calculation, the automatic unbiased machine learning diagnostic system 116 will not ignore any biological indicator. As the calculation process proceeds, the automatic unbiased machine learning diagnostic system 116 pays more and more attention to a first subset of biological indicators related to a specific health classification, and pays less attention to a second subset of biological indicators that are not related to a specific health classification (e.g., noise components). Depending on the specific embodiment or other considerations, changes to the indicator settings for machine learning operations can be protected by a user identity authentication system to ensure that the operation is unbiased. Depending on the specific implementation or other considerations, machine learning is deep learning, neural network, linear discriminant analysis, quadratic discriminant analysis, support vector machine, random tree, nearest neighbor or a combination thereof.

[0060] In one embodiment, the automatic unbiased machine learning diagnosis system 116 compares the abundance or absence of the determined biomarkers associated with the health classification with the quantitative results of the corresponding biological indicators obtained from the subject to diagnose the state of the health classification of the subject (positive or negative). For example, when the quantitative results of the biological indicators obtained from the subject fall within a specific range of the biomarkers, it can be determined that the subject suffers from a certain disease.

[0061] In a specific embodiment, the automatic unbiased machine learning diagnosis system 116 determines the medical treatment effect of the disease by comparing the quantitative results of biomarkers obtained from subjects with the disease but not receiving treatment, subjects with the disease and receiving treatment, and healthy subjects who do not suffer from the disease (and do not receive treatment). Here, the medical treatment program may include, but is not limited to, exercise therapy, dietary supplements, weight loss, surgical intervention, device implantation, and therapeutic or preventive programs applicable to subjects who diagnose or identify health conditions. For example, when the quantitative results of biomarkers obtained from treated subjects are closer to the quantitative results of biomarkers obtained from healthy subjects compared to the quantitative results of biomarkers obtained from untreated subjects, it can be determined whether the medical measures have a medically beneficial effect on the treatment of the health condition. In a specific embodiment, the automatic unbiased machine learning diagnosis system 116 can also determine the progress of medical measures by comparing the quantitative results of biological indicators obtained from subjects with a certain health classification but not receiving treatment, subjects with the health classification and receiving treatment, and subjects who do not belong to the health classification (and do not receive treatment). For example, when the quantitative results of the biomarkers obtained from the treated subjects approximately match the quantitative results of the biomarkers obtained from the healthy subjects, it can be determined that the treatment can be terminated. In a specific embodiment, the automatic unbiased machine learning diagnosis system 116 can also determine the progress of the healthy classification in a manner similar to determining the progress of the treatment. In a specific embodiment, the automatic unbiased machine learning diagnosis system 116 can also determine or select an effective treatment from a variety of possible treatments by comparing the progress of various possible treatment measures.

[0062] exist Figure 1A In the example of , the diagnosis result display system 118 is connected to the computer-readable medium 102. The diagnosis result display system 118 represents a system controlled by an entity with a platform (e.g., a network service provider), which is capable of displaying biological indicators measured by the automatic unbiased machine learning diagnosis system 116, and / or diagnostic results generated by the automatic unbiased machine learning diagnosis system 116. The entity may be the same as or different from the entity controlling the glycomics indicator quantification system 104, the genomics indicator quantification system 106, the proteomics indicator quantification system 108, the metabolomics quantification system 110, the lipidomics quantification system 112, the clinical indicator generation system 114, and / or the automatic unbiased machine learning diagnosis system 116.

[0063] Non-limiting examples of available platforms include, but are not limited to, web pages (for example, the measured biological indicators and / or diagnostic results can be displayed as messages on a personal web page, such as a personal web page of a hospital), electronic messages (such as emails, text messages, voice messages), print media (such as letters), and other platforms suitable for providing content to subjects.

[0064] Figure 1A A specific operation example of using the system to determine biological indicators for specific health classifications and diagnosing subjects based on the biological indicators is shown, which will be described below. The glycomic indicator quantification system 104 quantifies glycomic indicators (e.g., N-glycans) in a biological sample (e.g., a blood sample) and provides relevant information about the glycomic indicator quantification results to the automatic unbiased machine learning diagnosis system 116. Similar to the glycomic indicator quantification system 104, the genomic indicator quantification system 106, the proteomics indicator quantification system 108, the metabolomics indicator quantification system 110, and the lipidomics indicator quantification system 112 quantify the corresponding biological indicators in the biological sample, respectively, and provide relevant information about the quantitative results to the automatic unbiased machine learning diagnosis system 116. The clinical indicator generation system 114 generates clinical indicators of biological samples (e.g., positive / negative evaluations made on the questionnaire of each subject) and provides relevant information about the clinical indicators to the automatic unbiased machine learning diagnosis system 116.

[0065] Based on at least one quantitative result of the glycomic index received from the glycomic index quantification system 104, the genomic index received from the genomic index quantification system 106, the proteomics index received from the proteomics index quantification system 108, the metabolic index received from the metabolic index quantification system 110, and the lipidomics index received from the lipidomics index quantification system 112, and / or based on the quantitative and / or non-quantitative results of the clinical index received from the clinical index generation system 114, the automatic unbiased machine learning diagnosis system 116 can determine one or more biological indicators associated with one or more health classifications. The advantage is that the automatic unbiased machine learning diagnosis system 116 can determine one or more biological indicators as biomarkers based on a combination of data from two or more of the glycomic index quantification system 104, the genomic index quantification system 106, the proteomics index quantification system 108, the metabolic index quantification system 110, the lipidomics index quantification system 112, and the clinical index generation system 114, thereby improving the accuracy of the biological indicators as biomarkers.

[0066] In a specific embodiment, the automatic unbiased machine learning diagnosis system 116 diagnoses the subject based on the comparison of the biological indicators with the measured values ​​or the examination status of the subject. The diagnosis result display system 118 is used to display (e.g., generate a GUI) the biological indicators measured by the automatic unbiased machine learning diagnosis system 116 and / or the diagnosis results (e.g., positive or negative) generated by the automatic unbiased machine learning diagnosis system 116.

[0067] In order to quantify biological indicators (e.g., glycomic indicators, genomic indicators, proteomic indicators, metabolomics indicators, lipidomics indicators) respectively, system 100 can perform one or more quantitative operations on the unified mass spectrometry data obtained using the mass spectrometry technology used in a specific embodiment of the present invention. In some embodiments, for example, one or more peak picking tools and related integration methods can be used to quantify one or more corresponding biological indicators of a biological sample set or a single biological sample. In some embodiments, systems such as system 100 in the present invention can be equipped with subsystems or platforms, and one or more of systems 104-112 can use the subsystems or platforms when performing quantitative operations. Figure 1B An exemplary implementation of such an embodiment is shown.

[0068] Figure 1B Schematic diagram of an example system, according to one or more embodiments of the present patent, the system uses a peak integration platform to quantify biological indicators, identify one or more biological indicators associated with one or more health classifications, and predictively diagnose one or more health states of one or more subjects based on the biological indicators.

[0069] like Figure 1B As shown, the system 120 may include Figure 1A One or more of the components 102-118 mentioned in the figure can effectively communicate with the peak integration platform 130, the sample data repository 122, the transition list repository 124 and the glycoproteomics universe repository 126. As shown in the figure, the peak integration platform can be equipped with one or more acquisition components 132, feature extraction components 134, consensus / ensemble components 136 and peak integration components 138.

[0070] The acquisition component 132 can acquire a mass spectrometry dataset from a data source (e.g., a sample data repository 122) and allow one or more other components / parts of the system 120 to access the mass spectrometry dataset information, including, for example, one or more peak integration platforms 130, such as a feature extraction component 134, a consensus / integration component 136, and a peak integration component 138. The acquisition component 132 can also store a copy of the acquired dataset in one or more other data repositories connected thereto. The acquisition component 132 can acquire data in response to a command issued by a user, or automatically trigger the acquisition of data (e.g., a preset or periodic data extraction from a specific data source at a specific time), or continuously acquire data. For example, the acquisition component 132 can receive a user instruction (e.g., a user selects via a computing device) that the user desires to load a specific mass spectrometry dataset associated with a new biological sample from a subject being studied. The acquisition component 132 can also cause one or more components to acquire the dataset sequentially, simultaneously (i.e., in parallel), according to a preset order, or in another arrangement based on a preset criterion. The acquisition component 132 may be a stand-alone application that facilitates downloading mass spectrometry data set information in a dedicated manner, or it may operate in conjunction with another application to accomplish the same functionality.

[0071] The feature extraction component 134 can receive mass spectrometry data (e.g., associated with one or more biological samples from one or more subjects) from the acquisition component 132 and extract (i.e., identify) one or more representative proteomics features within the data. To achieve feature extraction, the feature extraction component can extract peptide-induced signals (i.e., peaks) from raw mass spectrometry data or preprocessed mass spectrometry data. The mass spectrometry data set associated with the biological sample from the subject can contain tens to thousands of spectra (corresponding to the intensity information of many different mass channels corresponding to isotopes) associated with many different molecular species (e.g., different molecules). The feature extraction component 134 can analyze the mass spectrometry data set to determine whether any spectra observed in the data set (e.g., observed isotope distributions, peaks, etc.) correspond to known or unknown but statistically important / significant molecular species. Known spectra and / or isotope distributions corresponding to known molecular species can be stored in the transition list repository 124, which the feature extraction component 134 can access during operation. For example, the transition list repository 124 can include known information related to transitions between peaks and troughs and related to a certain feature. The transition list repository 124 may also include preset peak waveforms with preset integration start and end points (the start and end points typically correspond to the valleys on either side of the peak associated with the known feature). Because mass spectrometry data is typically cluttered with overlapping isotope patterns and a large amount of interference information, it is necessary for the feature extraction component 134 to be able to identify combinations formed by overlapping peaks and filter out or otherwise reduce chemical and / or detector interference information in the data set.

[0072] The feature extraction component 134 uses peak picking tools known in the art, such as NITPICK, Skyline, OpenMS, DIA-Umpire, PECAN, XCMS, multiplierz, MZmine, T-BioInfo, MASS++, msInspect, MassSpecWavelet, MALDIquant, EigenMS, PrepMS, LC-IMS-MS-Feature-Finder, mMass, IMTBX (Ion Mobility Toolbox), Grppr (Grouper), mzDesktop, Cromwell, MapQuant, pParse, MzJava, HappyTools, Mass-UP, LIMPIC, SpiceHit, ProteinPilot, PROcess, GAGfinder, Intact Mass, JUMBO, Maltcms, SpectroDive, enviPick, findMF, PNNL PreProcessor, msXpertSuite, LCMS-2D or Siren (Sparse Isotope RegressionN). The feature extraction component 134 may only apply or enable any one or more of the aforementioned unbiased features, thereby prohibiting human intervention in the peak picking process.

[0073] In some embodiments, the feature extraction component may perform two or more peak picking operations on a data set (e.g., in parallel) to obtain two or more feature extraction result sets for the data set. The consensus / integration component 136 obtains multiple sets of feature extraction data from the data set of the feature extraction component 134, and identifies consensus or non-consensus between the multiple sets of feature extraction results or between the parts of the multiple sets of feature extraction results. Consensus can be considered based on features in the entire data set, or it can be considered based on other desired criteria. In some embodiments, consensus for a particular extracted feature (i.e., for a given peak (and transition related)) can be achieved by a preset value, percentage, or ratio of the peak picking operation applied to identify the peak in the data set.

[0074] In some embodiments, the consensus / integration component 136 can generate a consensus data set, the consensus data set includes a set of feature extraction results, the feature extraction results include feature extraction data; based on these feature extraction data, consensus is obtained through multiple peak picking operations. In some embodiments, multiple peak picking operations are performed on the extracted features, and the results are substantially similar. The consensus / integration component 136 can generate an integrated data set, the integrated data set includes a set of feature extraction results that can represent the extracted features. In some embodiments, one or more features of the peak (such as troughs, transitions, peaks, the slope of the peak waveform at a point on the waveform, etc.) are defined by statistical operations, and the feature extraction results in a set of multiple extracted feature results (for example, based on specific features) are merged, and the consensus / integration component 136 generates a collection data set. Such statistical operations may include one or more of the average, median, weighted combination, or any other combination.

[0075] The peak integration component 138 is capable of taking one or more feature extraction results from one or more feature extraction components 134 and consensus / integration components 136 (or other components or elements of the system 120) and integrating them to determine the area under the intensity curve that defines the peak associated with a given extracted feature (e.g., a given molecule). The peak integration component 138 can use any type of integration method, such as trapezoidal integration, rectangular integration, etc. The area under the intensity curve of a given feature (even a unit-free area) corresponds to the number of molecules associated with the feature within the considered range in the biological sample. Although the system of the present patent does not require the generation of a graph or graphic table of a spectrum, peak waveform, or any other data to operate, Figure 1C , 1D Figures 1 and 1E provide example diagrams that illustrate these concepts.

[0076] Figure 1CAn example of mass spectral data acquired by acquisition component 132 is shown. Feature extraction component 134 can identify these spectra as being associated with different features. For example, feature extraction component 134 can determine that the spectra generally marked by number 141 (having substantially similar mass-to-charge ratios) are associated with a first feature (e.g., a first peak); feature extraction component 134 can determine that the spectra generally marked by number 142 (having substantially similar mass-to-charge ratios) are associated with a second feature (e.g., a second peak); feature extraction component 134 can determine that the spectra generally marked by number 143 (having substantially similar mass-to-charge ratios) are associated with a third feature (e.g., a third peak); feature extraction component 134 can determine that the spectra generally marked by number 144 (having substantially similar mass-to-charge ratios) are associated with a fourth feature (e.g., a fourth peak); feature extraction component 134 can determine that the spectra generally marked by number 145 (having substantially similar mass-to-charge ratios) are associated with a fifth feature (e.g., a fifth peak).

[0077] from Figure 1C As can be seen, the spectrum of the fourth peak 144 overlaps the spectrum of the fifth peak 145. The spectrum of peak 144 is represented by a dashed line to distinguish it from the spectrum of the fifth peak 145. As described above, the feature extraction component 134 is able to distinguish between the two waveforms and identify them as spectra representing two different features, not the same feature. For the purpose of illustration, although Figure 1C Only two features are shown, and it should be understood that the feature extraction component is capable of and / or is trained to distinguish more than two overlapping peaks, and in particular is capable of determining or otherwise identifying the transition points between peaks and troughs associated with different features (to identify the starting and end points for subsequent integration).

[0078] Figure 1D Shown by Figure 1C , ... Figure 1D The first peak waveform 151 corresponds to Figure 1C The first peak in the middle is 141; similarly, Figure 1D The second peak, third peak, fourth peak and fifth peak waveforms marked with numbers 152, 153, 154 and 155 correspond to the following waveforms: Figure 1C The numbers in the figure are 142, 143, 144, 145 respectively for the second peak, the third peak, the fourth peak and the fifth peak.

[0079] Figure 1E Shows Figure 1D The example peak waveform shown in FIG. The area below the peak waveform curve is shaded to symbolically illustrate an example of the integration performed by the peak integration component 138. As shown in the figure, Figure 1BThe system 120 in the figure determines the starting point and the end point of the integration along the horizontal axis. For example, the system 120 can determine the transition point corresponding to 154a on the horizontal axis as the starting point for integrating the peak waveform 154, and the transition point corresponding to 154b on the horizontal axis should be used as the end point for integrating the peak waveform 154. Similarly, as shown in the figure, the system 120 can determine the transition point corresponding to 155a on the horizontal axis as the starting point for integrating the peak waveform 155, and the transition point corresponding to 155b on the horizontal axis as the end point for integrating the peak waveform 155.

[0080] Figure 2 A flowchart 200 of an example method is depicted, the method determining one or more biological indicators as one or more biomarkers associated with one or more health classifications, and diagnosing a subject based on the determined biomarkers. Flowchart 200 and other flowcharts in this specification are described in the form of a series of modules. It should be understood that the order of the modules can be changed and the modules can be rearranged for serial or parallel processing if necessary.

[0081] like Figure 2 As shown, the flowchart 200 begins with module 202: obtaining a quantitative result of at least one biological indicator. In a specific embodiment, the biological indicator is obtained by analyzing a biological sample. The biological indicator may include, for example, a glycomic indicator, a genomic indicator, a proteomic indicator, a metabolomics indicator, and a lipidomics indicator.

[0082] like Figure 2 As shown, the flowchart 200 proceeds to block 204: obtaining quantitative and / or non-quantitative results of clinical indicators. In a specific embodiment, the results and indicators are obtained by examining and questioning the subject.

[0083] like Figure 2 As shown, the flowchart 200 proceeds to block 206: performing an automatic unbiased machine learning operation to determine one or more biological indicators as one or more biomarkers for health classification. In the implementation process, the automatic unbiased machine learning operation starts with equal treatment of biological and clinical indicators, eliminates scientific bias, and does not configure the user to manually change the calculation settings of the machine learning operation.

[0084] like Figure 2As shown, flowchart 200 proceeds to module 208: diagnosing the health classification status (e.g., positive or negative) of the subject based on the comparison of the biological indicators obtained from the biological sample of the subject with the determined biomarkers. For example, when the abundance (e.g., above a threshold) of N-glycans and immunoglobulin G (IgG) obtained from the serum is determined to be a biomarker of ovarian cancer, the determination is made by whether the abundance of the corresponding biological indicators (i.e., N-glycans and IgG) obtained from the serum of the subject is sufficient (e.g., above a threshold). Module 208 is an optional module.

[0085] like Figure 2 As shown, flowchart 200 ends at module 210: If a determined biomarker and / or diagnostic result is obtained through module 208, it is displayed. In one embodiment, the diagnostic result is displayed by displaying the result on a web page, notifying the result by email, and / or inviting the result to be displayed on the medical device.

[0086] Figure 3 An example diagram 300 of a system for performing an automated unbiased deep learning operation to determine a biological indicator that can predict a subject classification, or alternatively, a classification based on a candidate biological indicator is shown. Diagram 300 includes a quantitative results data repository 301, a data classification engine 302, a training data set data repository 303, a test data set data repository 304, an unbiased deep learning engine 305, an internal validation engine 306, a new results input engine 307, and an external validation engine 308.

[0087] like Figure 3 As shown, the quantitative result data repository 301 represents the quantitative results obtained by digitizing the biological sample, and no matter what format is used, it must be compatible with the subsequent processing flow to determine the candidate biological indicators of the biomarker. More specifically, for example, when the glycomic index is quantified, the data unit of the quantitative result is associated with the unique identifier of the biological sample (or subject), and includes quantitative results of different types of glycosylated peptide fragments (e.g., known peptide fragments and / or unknown peptide fragments), and the index of the glycosylated peptide fragment is associated with the health classification status (e.g., positive / negative) representing one or more health classifications that each subject has or does not have.

[0088] like Figure 3As shown, the data classification engine 302 is connected to the quantitative result data storage repository 301. The data classification engine 302 represents dedicated hardware and software, which divides the quantitative results in the quantitative result data storage area 301 into two different data groups, namely a training data group and a test data group; the training data group is used to determine the candidate biological indicators through automatic unbiased deep learning, and the test data group is used to verify the determined candidate biological indicators. The manner of classifying each data unit into one of the training and test data groups, as well as the ratio of the training data group to the test data group (ratio of the training group to the test group), is not particularly limited, and a variety of data classification schemes can be adopted according to the algorithm.

[0089] like Figure 3 As shown, the training data set data repository 303 is connected to the data classification engine 302. The training data set data repository 303 represents data units classified as training data sets by the data classification engine 302. The data formats of the data units in the training data set data repository 303 and the quantitative result data repository 301 may be the same or different. In implementation, the data units in the quantitative result data set data repository 301 may be in an unstructured data format, and the data units in the training data set data storage area 303 may be in a structured data format.

[0090] like Figure 3 As shown, the test data group data repository 304 is connected to the data classification engine 302. The test data group data repository 304 represents data units classified as the test data group by the data classification engine 302. Similar to the training data group data repository 303, the data formats of the data units in the test data group data repository 304 and the data units in the quantitative result data repository 301 may be the same or different. In implementation, the data units in the quantitative result data repository 301 may be in an unstructured data format, and the data units in the test data group data repository 304 may be in a structured data format.

[0091] like Figure 3 As shown, the unbiased deep learning engine 305 is connected to the training data set data repository 303. The unbiased deep learning engine 305 represents dedicated hardware and software that performs an unbiased deep learning process according to an algorithm to determine one or more biological indicators as one or more candidate biomarkers indicating a classification (e.g., a disease state) of a subject.

[0092] In implementation, the unbiased deep learning engine 305 forms an artificial neural network (ANN), which includes an input layer, an output layer, and one or more hidden layers formed between the input layer and the output layer. The input layer includes a plurality of artificial neurons, and for each artificial neuron in the input layer, a quantitative result of some or all types of glycosylated peptide fragments is input; as an alternative, one or more indicators representing the state of the subject can also be input. Similarly, each hidden layer in the one or more hidden layers contains a plurality of artificial neurons, and for each artificial neuron in each hidden layer in the one or more hidden layers, the output of one or more artificial neurons in the directly adjacent previous layer (e.g., one of the input layer or hidden layers) is input. In each artificial neuron in the one or more hidden layers, the input of the directly adjacent previous layer is received with a certain weight according to the algorithm, and a specific calculation (e.g., XOR) is performed. The output of the artificial neuron of the last hidden layer in the one or more hidden layers is input to the one or more artificial neurons in the output layer, and the output layer outputs one or more biological indicators as candidate biomarkers to predict classification (e.g., disease state). Depending on the specific implementation or other considerations, the artificial neural network of the unbiased deep learning engine 305 may include a neural network, such as a feedforward neural network in which the connections between the layers do not form a loop, or a recurrent neural network (RNN) in which the connections between the layers form a directed loop. Depending on the specific implementation or other considerations, a single unit of the unbiased deep learning engine 305 can perform a deep learning process for multiple target health classifications. In an alternative, the unbiased deep learning engine 305 provides independent units for target health classifications.

[0093] like Figure 3As shown, the internal validation engine 306 is connected to the unbiased deep learning engine 305 and the test data set data repository 304. The output of the internal validation engine 306 is also connected to the data classification engine 302 and the unbiased deep learning engine 305. The internal validation engine 306 represents dedicated hardware and software that validates one or more candidate biological indicators determined by the unbiased deep learning engine 305 by matching the candidate biological indicators with data units of the test data set (in the test data set data repository 304), and outputs valid candidate biological indicators as biomarkers associated with health classification. In a specific embodiment, for each of the one or more candidate biological indicators, the internal validation engine 306 is used to determine whether the quantitative results of the candidate biological indicator obtained from the positive subjects in the test data set (i.e., subjects with a certain health classification) match the abundance (or absence) of the candidate biological indicator determined from the data units in the training data set, and determine whether the quantitative results of the candidate biological indicator obtained from the negative subjects in the test data set (i.e., subjects without a certain health classification) match the absence (or abundance) of the candidate biological indicator determined from the data units in the training data set.

[0094] In a specific implementation, the matching result obtained by the internal verification engine 306 is fed back to the data classification engine 302, and based on the matching result, the data classification engine 302 maintains or modifies the way of classifying the quantitative results into the training data group and the test data group. In a specific implementation, the matching result obtained by the internal verification engine 306 is fed back to the unbiased deep learning engine 305; based on the matching result, the unbiased deep learning engine 305 maintains or modifies the weight of each artificial neuron used in the artificial neural network.

[0095] like Figure 3 As shown, the new result input engine 307 is connected to the quantitative result data repository 301. The new result input engine 307 represents dedicated hardware and software, which is dedicated to inputting the quantitative results of biological indicators of one or more new subjects (or new biological samples) into the system. New subjects include, for example, subjects who are waiting for predictive diagnosis of health classification based on biomarkers, and / or subjects who have been diagnosed as having or not having a certain health classification. The quantitative results of the new subjects are input into the quantitative result data repository 301 as additional data units of the new subjects, and are input into the external verification engine 308. Based on the quantitative results of the new subjects, predictive diagnosis is performed on the new subjects or extended verification of the biomarkers is performed.

[0096] like Figure 3As shown, the external validation engine 308 is connected to the internal validation engine 306 and the new result input engine 307. The output of the external validation engine 308 is also connected to the data classification engine 302 and the unbiased deep learning engine 305. In a specific implementation, the external validation engine 308 represents dedicated hardware and software that performs predictive diagnosis based on one or more biomarkers validated by the internal validation engine 306; and / or performs extended validation of one or more biomarkers by matching valid biomarkers with data units of new subjects input from the new result input engine 307. In a specific implementation, for the purpose of predictive diagnosis, for each of the one or more biomarkers, the external validation engine 308 determines whether the quantitative results of the corresponding biological indicators obtained from the positive subjects match the abundance or absence of the biomarker. In another specific implementation, for the purpose of extended validation, for each of the one or more biomarkers, the external validation engine 308 determines whether the quantitative results of the biological indicators obtained from the positive subjects (i.e., subjects with a certain health classification) in the new subjects match the abundance or absence of the biomarker, and whether the quantitative results of the corresponding biological indicators obtained from the negative subjects (i.e., subjects without a certain health classification) in the new subjects match the absence of the biomarker. Then, for display purposes, the external validation engine 308 outputs the valid biomarkers.

[0097] In a specific embodiment, similar to the internal verification engine 306, the matching results obtained by the external verification engine 308 are fed back to the data classification engine 302; based on the matching results, the data classification engine 302 maintains or modifies the way the quantitative results are divided into the training data group and the test data group, and / or the ratio of training to testing. In addition, the external verification engine 308 feeds back the obtained matching results to the unbiased deep learning engine 305, and based on the matching results, the unbiased deep learning engine 305 maintains or modifies the weight of each artificial neuron applied by the artificial neural network, and / or other operational indicators of deep learning to improve the accuracy of determining the health classification.

[0098] Figure 4 An example flowchart 400 of the invention depicts a method for determining biomarkers for predicting a subject classification by performing an automated unbiased deep learning operation, and predicting a classification based on the determined biomarkers. Flowchart 400 begins at module 402, which is to classify quantitative results obtained by digitizing a biological sample into a training data set and a test data set.

[0099] like Figure 4As shown, the flowchart 400 proceeds to block 404: performing an unbiased deep learning process on the training data set to determine one or more biological indicators as one or more candidate biomarkers for predicting health classification.

[0100] like Figure 4 As shown, the flowchart 400 proceeds to block 406: verifying the determined candidate bio-indicators based on the test data set. In a specific embodiment, the verification includes determining whether a positive subject of a certain healthy classification has quantitative results of one or more bio-indicators that match the abundance or absence of the determined candidate bio-indicators, and determining whether a negative subject of the healthy classification has quantitative results that do not match the abundance or absence of the determined candidate bio-indicators.

[0101] like Figure 4 As shown, the flowchart 400 proceeds to decision point 408, that is, determining whether each of the one or more candidate biomarkers is valid. If there is an invalid candidate biomarker (408-N), the flowchart 400 will proceed to module 410, and the verification result of the candidate biomarker will be fed back to the quantitative result classification performed by module 402 and / or the deep learning performed in module 404, and the flowchart 400 ends. If the candidate biomarker is valid (408-Y), the flowchart proceeds to module 412, and the quantitative result classification performed by module 402 and / or the deep learning process performed in module 404 are fed back. In a specific embodiment, for an invalid candidate biomarker, the neural connection between the two artificial neurons can be weakened, for example, the weight of the invalid candidate biomarker can be reduced; for an effective candidate biomarker, the neural connection between the two artificial neurons can be strengthened, for example, the weight of the effective candidate biomarker can be increased.

[0102] like Figure 4 As shown, the flowchart 400 proceeds to decision point 414: determining whether to perform a predictive diagnosis of the health classification of the new subject. If it is determined that a predictive diagnosis of the health classification of the new subject is performed (414-Y), that is, if the health classification status of the new subject is unknown, the flowchart 400 proceeds to module 416, and a predictive diagnosis is performed on the new subject based on the comparison between the abundance or absence of the effective biomarker (verified in module 406) and the quantitative results of the corresponding biological indicators obtained from the biological sample of the new subject, and the flowchart 400 ends. For example, when the abundance of a certain glycosylated peptide fragment exceeds a preset threshold, indicating that the health classification status is positive, then when the quantitative results of the glycosylated peptide fragment obtained from the biological sample of the new subject are obtained, the health classification status of the new subject is determined to be positive. In a specific embodiment, the invalid biomarker (verified invalid in module 406) is not used for the predictive diagnosis in module 416.

[0103] On the other hand, if it is determined that a predictive diagnosis of a health classification is not to be performed on the new subject (414-N), for example, if the health classification status of the new subject is known, then the flowchart 400 proceeds to module 418 to perform an extended validation of the effective biomarkers based on the quantitative results of the new subject. In a specific embodiment, the extended validation includes determining whether a positive healthy classification subject has quantitative results of one or more corresponding biological indicators that match the abundance or absence of the effective biomarkers, and determining whether a negative healthy classification subject has quantitative results of one or more corresponding biological indicators that do not match the abundance or absence of the effective biomarkers.

[0104] like Figure 4 As shown, flowchart 400 proceeds to decision point 420 to determine whether each of the one or more biomarkers has passed extended validation. If there are invalid biomarkers (420-N), flowchart 400 returns to module 410 and continues as described above. If there are biomarkers that have passed extended validation (420-Y), flowchart 400 proceeds to module 422, and the quantitative result classification performed in module 402 and / or the deep learning process performed by module 404 are fed back in a manner similar to module 412, and flowchart 400 ends. In a specific embodiment, for invalid biomarkers, the neural connection between the two artificial neurons can be weakened, such as reducing the weight of the invalid biomarker. For biomarkers that have been extended and validated, the neural connection between the two artificial neurons can be strengthened, such as further increasing the weight of the biomarker that has been extended and validated.

[0105] Figure 5 500 is an exemplary system diagram for diagnosing a health condition of a subject based on a machine learning process and quantitative results of corresponding biological indicators of the subject obtained from a biological sample of the subject. Diagram 500 includes a standard biomarker data repository 501, a quantitative result data repository 502, a biomarker diagnosis engine 503, and a diagnosis result data repository 504.

[0106] like Figure 5 As shown, the standard biomarker data repository 501 represents detailed information of biomarkers determined by an automated unbiased machine learning process, which can be obtained from, for example, Figure 3The internal validation engine 306 and / or the external validation engine 308 shown are obtained. The detailed information of the biomarker includes, for example, N-glycans obtained from serum above a first threshold and IgG above a second threshold, indicating that ovarian cancer is positive. In another example, the detailed information of the biomarker includes a glycosylated peptide fragment above a certain threshold and a blood glucose level below a certain threshold, indicating that cancer is positive. As described above, any single biological indicator or a combination of two or more biological indicators can be used as a biomarker.

[0107] like Figure 5 As shown, the quantitative result data repository 502 represents the quantitative results of quantifiable biological indicators and data of non-quantifiable biological indicators from the biological samples of the subjects. In implementation, the quantitative results and data are from Figure 1A One or more of the glycomics index quantification system 104 , the genomics index quantification system 106 , the proteomics index quantification system 108 , the metabolomics index quantification system 110 , the lipidomics index quantification system 112 , and the clinical index generation system 114 are shown.

[0108] like Figure 5 As shown, the biomarker diagnosis engine 503 is connected to the standard biomarker data repository 501 and the quantitative result data repository 502. In a specific embodiment, the biomarker diagnosis engine 503 represents dedicated hardware and software, which diagnoses the subject according to one or more biomarkers and stores the diagnosis results in the diagnosis result data repository 504. In a specific embodiment, the biomarker diagnosis engine 503 determines whether the quantitative results of the biological indicators obtained from the biological sample of the subject are within a specific range based on the biomarkers, and / or whether the non-quantitative data of the non-quantitative indicators obtained from the subject matches the standards of the biomarkers, thereby determining whether the subject has a certain health classification.

[0109] In a specific embodiment, based on the detailed information of the biomarker, the biomarker diagnostic engine 503 determines whether the quantitative results of the biological indicators obtained from the biological sample of the subject are close to a specific range corresponding to the health status, or whether the quantitative results deviate from another specific range corresponding to the health classification status, and by comparing with the quantitative results before the subject receives treatment, it can be determined whether the treatment measures applied to the subject are effective.

[0110] In a specific embodiment, according to the detailed information display of the biomarker, the biomarker diagnosis engine 503 determines whether the quantitative results of the biological indicators obtained from the biological sample of the subject increase or decrease within a specific range related to the health classification state, and whether the quantitative results deviate from another specific range related to the health state, and by comparing the previous quantitative results of the subject after being diagnosed with the health classification, so as to determine the objective health classification progress of the subject. For example, after the subject is diagnosed with heart disease, the stage of the heart disease is objectively determined based on the level of the biomarker.

[0111] In a specific embodiment, the biomarker diagnosis engine 503 determines (or selects) a treatment method suitable for a subject with a certain health classification based on the diagnosis results, especially the treatment effectiveness results, stored in the diagnosis result data repository 504. For example, based on the quantitative results and biomarkers of the subject, the biomarker diagnosis engine 503 retrieves the treatment effects of multiple different treatment methods that have been applied to subjects with the health classification from the diagnosis result database 504, and selects the best treatment method from the multiple treatment methods.

[0112] disease

[0113] The patented method is applicable to any disease or symptom that can be detected by analyzing biological indicators obtained from a biological sample of a subject. In some embodiments, the disease or symptom is cancer. In other embodiments, the cancer is acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), adrenocortical carcinoma, anal cancer, bladder cancer, blood cancer, bone cancer, brain tumor, breast cancer, female reproductive system cancer, male reproductive system cancer, central nervous system lymphoma, cervical cancer, childhood rhabdomyosarcoma, childhood sarcoma, chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), colon and rectal cancer, colon cancer, endometrial cancer, endometrial sarcoma, esophageal cancer, eye cancer, gallbladder cancer, gastric cancer, gastrointestinal cancer, hairy cell leukemia, head and neck cancer, hepatocellular carcinoma, Hodgkin's disease, hypopharyngeal cancer, Kaposi's sarcoma, kidney cancer, laryngeal cancer, Leukemia, liver cancer, lung cancer, malignant tumor fibrous histiocytoma, malignant thymoma, melanoma, mesothelioma, multiple myeloma, myeloma, nasal and paranasal sinus cancer, nasopharyngeal cancer, nervous system cancer, neuroblastoma, non-Hodgkin's lymphoma, oral cancer, oropharyngeal cancer, osteosarcoma, ovarian cancer, pancreatic cancer, parathyroid cancer, penile cancer, pharyngeal cancer, pituitary tumor, plasma cell tumor, primary central nervous system lymphoma, prostate cancer, rectal cancer, respiratory system cancer, retinoblastoma, salivary gland cancer, skin cancer, small intestine cancer, soft tissue sarcoma, stomach cancer, testicular cancer, thyroid cancer, urinary system cancer, uterine sarcoma, vaginal cancer, vascular system, macroglobulinemia, Wilms tumor, etc. In another embodiment, the cancer is breast cancer, cervical cancer or ovarian cancer.

[0114] In another embodiment, the disease is an autoimmune disease. In another embodiment, the autoimmune disease is acute disseminated encephalomyelitis, Addison's disease, aglobulinemia, age-related macular degeneration, alopecia areata, amyotrophic lateral sclerosis, ankylosing spondylitis, antiphospholipid syndrome, antisynthetase syndrome, atopic allergy, atopic dermatitis, autoimmune autoimmune diseases cardiomyopathy, autoimmune intestinal disease, autoimmune hemolytic anemia, autoimmune hepatitis, autoimmune inner ear disease, autoimmune lymphoproliferative syndrome, autoimmune peripheral neuropathy, autoimmune pancreatitis, autoimmune polyendocrine syndrome, autoimmune progestogen dermatitis, autoimmune thrombocytopenic purpura, autoimmune urticaria concentric sclerosis, Behcet's disease, Burghardt's disease, Jervis disease, Bickerstaff encephalitis, Blau syndrome, bullous pemphigus, cancer, Castleman disease, celiac disease, Chagas disease, chronic inflammatory demyelinating polyneuropathy, chronic recurrent focal osteomyelitis, chronic obstructive pulmonary disease, allergic granulomatous vasculitis, cicatricial pemphigus, Kogan syndrome, cold agglutinin disease, complement component 2 deficiency, contact dermatitis, cranial arteritis, CREST syndrome, Crohn's disease, Cushing's syndrome, cutaneous leukocytic vasculitis, DeGou's disease skin disease, dermatitis herpetiformis, dermatomyositis, type I diabetes, diffuse cutaneous systemic sclerosis, Deller syndrome, drug-induced lupus, discoid lupus erythematosus, eczema, endometriosis, arthritis associated with inflammatory inflammation, eosinophilic fasciitis, Eosinophilic enteritis, eosinophilia nodularis, fetal fibroblastosis, essential mixed cryoglobulinemia, Evan's syndrome, progressive osteoproliferative fibrosis, fibrosing pneumonia, gastritis, gastrointestinal pemphigus, glomerulonephritis, Goodpass syndrome, Graves' disease, Guillain-Barré syndrome, Hashimoto's encephalopathy, Hashimoto's thyroiditis, Henoch-Schönlein purpura, HIV, pemphigus gestationis, hidradenitis suppurativa, Hughes-Stowan syndrome, hypoglobulinemia, idiopathic inflammatory demyelinating disease, idiopathic pulmonary fibrosis, idiopathic thrombocytopenic purpura chronic, IgA nephropathy, inclusion body myositis, arthritis, Kawasaki disease, Lambert-Eaton myasthenic syndrome, leukocytoclastic vasculitis, lichen planus, planus Lichen, linear IgA disease, lupus erythematosus, Majeed syndrome, Meniere's disease, microscopic polyangiitis, mixed connective tissue disease, morphea, multiple Mann's disease, myasthenia gravis, myositis, narcolepsy, neuromyelitis optica, neuromyotonia, ocular cicatricial pemphigus, optic nerve innervation myoclonus syndrome, Odd's thyroiditis, gyral rheumatism, streptococcal autoimmune neuropsychiatric disease, parasympathetic degeneration, paroxysmal nocturnal hemoglobinuria, Parry-Romberg syndrome, Parsonage-Turner syndrome, Parspinitis, pemphigus vulgaris, pernicious anemia, venous encephalomyelitis, POEMS syndrome, polyarteritis nodosa, polymyalgia rheumatica, polymyositis,Primary biliary sclerosis, primary cholecystitis, psoriatic arthritis, pyoderma gangrenosum, pure red cell dysplasia, Rasmussen's encephalitis, Raynaud's syndrome, relapsing chondritis, Reiter's syndrome, restless legs syndrome, retroperitoneal fibrosis, rheumatoid arthritis, rheumatic fever, sarcoidosis, schizophrenia, Schmidt's syndrome, Schnitzler's syndrome, scleritis, scleroderma, serum sickness, Sjögren's syndrome, spondyloarthritis, stiff-man syndrome, subacute bacterial endocarditis, Susac's syndrome, acute febrile neutrophilic dermatosis, sympathetic ophthalmia, Takayasu's arteritis, temporal arteritis, thrombocytopenia, painful ophthalmoplegia syndrome, transverse myelitis, ulcerative colitis, undifferentiated connective tissue disease, urticarial vasculitis, vasculitis, vitiligo and Wegener's disease, etc. In another embodiment, the autoimmune disease is AIDS, primary sclerosing cholangitis, primary biliary cirrhosis or psoriasis. Example

[0115] Example 1: Quantification of IgG glycopeptides as breast cancer biomarkers

[0116] Figure 6 The quantitative results of the changes in IgG1, IgG0, and IgG2 glycopeptides in the plasma samples of breast cancer patients relative to those of the control group are shown. The IgG1, IgG0, and IgG2 glycopeptides in the plasma samples of breast cancer patients at different cancer stages were analyzed relative to those of the age-corresponding control group, and their proportional changes were compared. Specifically, MRM quantitative analysis was performed on a QQQ mass spectrometer on 20 samples at the carcinoma in situ stage, 50 samples at the EC1 stage, 138 samples at the EC2 stage, 25 samples at the EC3 stage, 9 samples at the EC4 stage, and 73 age-matched control samples. The quantitative results are shown in Figure 2. Figure 6 As shown, in each stage of breast cancer studied in this experiment, the levels of some IgG1 glycopeptides were increased and the levels of some IgG1 glycopeptides were decreased compared to the control group. For example, in each stage of breast cancer studied in this experiment, Al-A11 among the IgG1 glycopeptides was monitored; compared to the control group, the levels of glycopeptides A1 and A2 were found to be increased, and the levels of glycopeptides A8, A9, and A10 were found to be decreased. Therefore, glycopeptides A1, A2, A8, A9, and A10 can be confirmed as biomarkers for breast cancer. It can be seen that the level of A5 increased, although the increase was small, compared to the control group; the level of A6 decreased, although the decrease was small, compared to the control group. Therefore, if "small" changes are considered sufficient, A5 and A6 can also be confirmed as biomarkers.

[0117] Example 2: Quantitative analysis of IgG glycopeptides as potential biomarkers for PSC and PBC

[0118] Example 2 shows the quantitative results of changes in IgG, IgM and IgA glycopeptides in plasma samples from patients with primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC) and healthy donors (not suffering from PBS or PSC), such as Figure 7 shown.

[0119] In Example 2, IgG1 and IgG2 glycopeptides in plasma samples of PSC patients, PBC patients, and plasma samples of healthy donors were analyzed, and the changes in their glycopeptide ratios were compared. Specifically, 100 PBC plasma samples, 76 PSC plasma samples, and 49 healthy donor plasma samples were subjected to MRM quantitative analysis on a QQQ mass spectrometer. Figure 7 The quantitative results show that in the plasma samples of PBC and PSC patients, some IgG1 glycopeptides are increased, while some IgG1 glycopeptides are decreased compared with healthy donors. For example, glycopeptide A is increased, while glycopeptides H, I, and J are decreased in the plasma samples of PBC and PSC patients compared with healthy donors. Therefore, glycopeptides A, H, I, and J can be confirmed as biomarkers for PBC and PSC.

[0120] also, Figures 8A-8C and Fig. 9 The results of individual and combined discriminant analysis (using the K-means clustering algorithm) are shown, indicating that the accuracy of predicting disease status in the combined discriminant analysis was 88%. Similar analyses were performed on IgA and IgM glycoproteins in plasma samples from PBC patients and PSC patients. Fig. 8A -C shows that the prediction accuracy based on the single data of IgG, IgM and IgA is 59%, 69% and 74% respectively. Fig. 9 As shown, after combining all the results for IgG, IgM, and IgA, the accuracy of the discriminant analysis was approximately 88%.

[0121] These and other examples provided in this specification are intended to illustrate, not to limit, the embodiments described. For example, the words "implementation" used in this specification refer to embodiments described by way of non-limiting example. The techniques described in the foregoing text and drawings may be mixed and matched as desired to produce other alternative embodiments.

Claims

1. A system, comprising: Memory; One or more processors, the one or more processors comprising: Glycomic index quantitative system, Genomics indicator quantification system, Proteomics index quantification system, Metabolic index quantification system, Lipidomics index quantitative system, and Clinical indicator generation system, and The one or more processors are electrically connected to a non-transitory computer-readable medium having machine-readable instructions stored thereon, which when executed by the one or more processors cause the system to: Quantifying one or more glycomic indices of one or more biological samples associated with one or more subjects; wherein the quantification comprises extracting one or more features using a glycomic indices quantification system in a mass spectrometry data set associated with the biological sample; determining a peak waveform associated with the one or more extracted features, and integrating the peak waveform; quantifying one or more other biological indicators selected from the group consisting of: genomic indicators, proteomic indicators, metabolomics indicators, and lipidomics indicators of one or more biological samples, wherein the genomic indicators are quantified using a genomic indicator quantification system, the proteomic indicators are quantified using a proteomic indicator quantification system, the metabolomics indicators are quantified using a metabolomics indicator quantification system, and the lipidomics indicators are quantified using a lipidomics quantification system; Using a clinical indicator generation system to obtain one or more clinical indicators, the one or more clinical indicators comprising quantifiable and / or non-quantifiable data obtained from a subject; Applying the quantified glycomic index, one or more of the quantified other biological indexes, and the one or more clinical indexes to an automated unbiased deep learning operation to identify one or more biomarkers; and A health status of the one or more subjects is determined based on a comparison of the one or more biomarkers to a threshold value. 2 . The system of claim 1 , wherein the quantification of the one or more glycomic indices comprises quantifying one or more glycosylated proteins.

3. The system of claim 2, wherein the one or more glycosylated proteins include one or more of the following: alpha-1-acid glycoprotein, alpha-1-antitrypsin, alpha-1B-glycoprotein, alpha-2-HS-glycoprotein, alpha-2-macroglobulin, antithrombin III, apolipoprotein B-100, apolipoprotein D, apolipoprotein F, beta-2-glycoprotein-1, ceruloplasmin, fetuin, fibrinogen, immunoglobulin (Ig) A, IgG, IgM, haptoglobin, heme, histidine-rich glycoprotein, kininogen-1, transferrin, vitronectin, and zinc-alpha-2-glycoprotein.

4. The system of claim 3, wherein the transferrin is serum transferrin. 5 . The system of claim 1 , wherein the quantification of the one or more glycomic indices comprises quantification of one or more glycosylated peptide fragments.

6. The system according to claim 5, wherein the one or more glycosylated peptide fragments include one or more of: O-glycosylated peptide fragments and N-glycosylated peptide fragments.

7. The system according to claim 5, wherein the one or more glycosylated peptide fragments have an average length of 5 to 50 amino acid residues.

8. The system of claim 5, wherein the glycosylated peptide fragments are generated by applying a fragmentation process to one or more glycosylated proteins.

9. The system of claim 8, wherein the fragmentation process comprises introducing at least two proteases into an environment containing one or more glycosylated proteins.

10. The system of claim 9, wherein the at least two proteases are selected from the group consisting of serine proteases, threonine proteases, cysteine ​​proteases, aspartic proteases, glutamic proteases, metalloproteases, and asparagine peptide cleaving enzymes; or wherein the at least two proteases are selected from the group consisting of trypsin, chymotrypsin, endoprotease Asp-N, endoprotease Arg-C, endoprotease Glu-C, endoprotease Lys-C, pepsin, thermolysin, esterase, papain, proteinase K, subtilisin, clostripain, and carboxypeptidase.

11. The system of claim 1, wherein quantification of the one or more glycomic indices comprises using coupled chromatography, wherein the coupled chromatography comprises performing a multiple reaction monitoring mass spectrometry (MRM-MS) operation.

12. The system of claim 11, wherein the coupled chromatography further comprises performing a liquid chromatography-mass spectrometry (LC-MS) operation.

13. The system of claim 1, wherein the quantifying one or more glycomic indices comprises employing one or more of triple quadrupole (QQQ) mass spectrometry and quadrupole time-of-flight (qTOF) mass spectrometry.

14. A system according to claim 1, wherein the one or more clinical indicators include one or more of the following: a heart rhythm measurement of a subject among the one or more subjects, a blood pressure measurement of the subject, a blood type measurement of the subject, a body temperature measurement of the subject, a skin color measurement of the subject, an eye color measurement of the subject, a blood glucose concentration measurement of the subject, a weight measurement of the subject, a height measurement of the subject, the subject's perceived health classification status or responses to questions.

15. The system of claim 1, further comprising: One or more biomarkers are compared to one or more of the following: (1) one or more of the glycomic indicators of one or more other subjects with the same health condition who have received treatment for the health condition, and (2) one or more of the glycomic indicators of one or more other subjects with the same health classification condition but who have not received treatment for the health condition, and (3) One or more of the glycomic indicators of one or more other subjects who do not have the same health classification status.

16. The system of claim 1, wherein the peak waveform associated with one or more extracted features is based on a mass spectrum.

17. The system of claim 1, wherein the integration of the peak waveform is performed using a transition list having preset start and end points.

18. The system of claim 1, wherein the one or more processors further comprises a diagnostic result delivery system configured to output one or more biomarkers and / or health conditions of one or more subjects.

Images