Methods and systems for protein identification

Abstract

This invention provides methods and systems for accurate and efficient protein identification and quantification. [Solution] A method for repeatedly identifying candidate proteins in a sample of an unknown protein is disclosed, comprising the steps of: receiving information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein, wherein each affinity reagent probe is configured to selectively bind to one or more candidate proteins; comparing at least a portion of the binding measurement information with a database containing a plurality of protein sequences, wherein each protein sequence corresponds to one candidate protein; and repeatedly generating the probability that each of the one or more candidate proteins is present in the sample, based on the comparison of the binding measurement information of the candidate proteins with a database containing a plurality of protein sequences.

Description

[Technical Field]

[0001] cross reference This application claims priority to U.S. Provisional Patent Application No. 62 / 575,976, filed on October 23, 2017. Yes, and the provisional application is incorporated herein by reference in its entirety. [Background technology]

[0002] background Current techniques for protein identification are typically highly specific and sensitive parental identification. Binding of a compatible reagent (such as an antibody) and subsequent reading of information, or from a mass spectrometer. It depends on the peptide reading data (typically about 12-30 amino acids long). Technologies like these enable highly specific and sensitive affinity reagents to target proteins of interest. Based on the analysis of binding measurements, the presence, absence, or quantity of the candidate protein is determined. Therefore, it can be applied to unknown proteins in the sample. [Overview of the project]

[0003] overview In this specification, improved identification and The need for quantification is recognized. The methods and systems provided herein involve the analysis of a sample. This can significantly reduce or eliminate errors in identifying proteins, and therefore This improves the quantification of the aforementioned protein. Such methods and systems are available for unknown proteins. This enables accurate and efficient identification of candidate proteins in protein samples. The identification is a parent that is set to selectively bind to one or more candidate proteins. This can be based on repeated calculations using information on the binding measurement of the symmetry reagent probe. In this manner, a sample of an unknown protein may be repeatedly exposed to individual affinity reagent probes, pooled affinity reagent probes, or combinations of individual affinity reagent probes and pooled affinity reagent probes. Identification may include an estimation of the confidence level that each of one or more candidate proteins is present in the sample. to pooled affinity reagent probes, or combinations of individual affinity reagent probes and pooled affinity reagent probes. Identification may include an estimation of the confidence level that each of one or more candidate proteins is present in the sample.

[0004] In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences. In one aspect, a computer-implemented method for repeatedly identifying each candidate protein in a sample of an unknown protein is disclosed herein, the method including the steps of: (a) receiving, by the computer, information on binding measurements for each of a plurality of affinity reagent probes for the unknown protein in the sample, each affinity reagent probe being set to selectively bind to one or more of a plurality of candidate proteins; (b) comparing, by the computer, at least a portion of the information on the binding measurements to a database including a plurality of protein sequences, each protein sequence corresponding to one of the plurality of candidate proteins; and (c) repeatedly generating, by the computer, for each of one or more of the plurality of candidate proteins, a probability that each of the one or more candidate proteins is present in the sample based on the comparison of the at least a portion of the information on the binding measurements for each of the one or more candidate proteins to the database including the plurality of protein sequences.

[0005] ​​In some embodiments, the step of generating the plurality of probabilities comprises a plurality of additional affinity reagents further repeatedly receiving additional information on the binding measurements for each of the probes where each of the additional affinity reagent probes is configured to selectively bind to one or more of the plurality of candidate proteins. In some embodiments, the method further comprises generating a confidence level for each of the one or more candidate proteins that the candidate protein matches one of the unknown proteins in the sample. In some embodiments, the step of generating the probabilities comprises taking into account

[0006] the error rate of the detector, which is related to the information on the binding measurements. In some embodiments, the error rate of the detector is obtained from the specification of one or more detectors used to obtain the information on the binding measurements. In some embodiments, the error rate of the detector is set to the estimated error rate of the detector. In some embodiments, the estimated error rate of the detector is set by the user of the computer. In some embodiments, the estimated error rate of the detector is about 0.001. Such an error rate may include the physical errors of the detector, as described elsewhere in this specification. Alternatively, such an error rate may be due to the failure of the probe to "land" on the protein, which may occur, for example, if the probe sticks in the system and is not properly washed away, or if the probe binds to a protein that was not expected based on previous characterization and testing of the probe. Thus, the error rate of the detector may include one or more of the following: the physical errors of the detector In some embodiments, the step of generating the probabilities comprises taking into account Error rate, off-target connection rate, or error rate due to stacked probes.

[0007] In some embodiments, the process of repeatedly generating the plurality of probabilities is carried out from subsequent iterations. Further removal of one or more candidate proteins from the aforementioned plurality of candidate proteins This includes several reactions necessary to carry out the repeated generation of the aforementioned probabilities. Reduces the amount. In some embodiments, remove one or more candidate proteins. This is done based at least on predetermined criteria for the binding measurement related to the candidate protein. In some embodiments, the predetermined criteria are the plurality of affinity reagent probes. The one or more weathers having a coupled measurement below a predetermined threshold for a first group among them Contains coproteins.

[0008] In some embodiments, each of the probabilities is relative to the length of the candidate protein. It is normalized. In some embodiments, each of the probabilities is the number of candidate proteins. The sum of the probabilities of the quality is normalized. In some embodiments, the multiple affinity tests The drug probe comprises 50 or fewer affinity reagent probes. In some embodiments, the multiple The affinity reagent probes include 100 or fewer affinity reagent probes. In some embodiments, In this configuration, the plurality of affinity reagent probes include 500 or fewer affinity reagent probes.

[0009] The length of the candidate protein is related to the binding of the candidate protein to a specific affinity reagent. Recognize that this is an approximate value that approximates the number of available epitopes ("binding sites") in the location. In addition, in some embodiments, each of the above probabilities is such that for each of the candidate proteins In this case, it is normalized to the total number of available binding sites. The number of available binding sites for each of the candidate proteins is determined using a qualitative process. Determined by experiment. In some embodiments, the qualitative process involves a specific protein The binding of the affinity reagent to the quality is measured repeatedly. In some embodiments, the qualitative Rothes during the aforementioned protein identification methods and systems described herein It will be carried out under conditions similar to or identical to those shown.

[0010] In some embodiments, the probability is generated repeatedly until a predetermined condition is met. In some embodiments, the predetermined conditions are met with a probability of at least 90% confidence. This includes generating each of the following. In some embodiments, the predetermined conditions are less This includes generating each of the aforementioned multiple probabilities with a 95% confidence level. In some embodiments, In this case, the predetermined conditions generate each of the plurality of probabilities with at least 99% confidence. This includes doing so.

[0011] In some embodiments, the method involves extracting one or more unknown proteins from the sample. The process further includes the step of generating a paper or electronic report to identify. In some embodiments, The sample includes a biological sample. In some embodiments, the biological sample is , obtained from the subject. In some embodiments, the method is based on at least the plurality of probabilities. The process further includes a step of identifying the disease state in the subject.

[0012] In some embodiments, the method counts the number of identifications made for each candidate protein. The procedure further includes the step of quantifying the protein in the biological sample. In that embodiment, the number of raw proteins is not limited, but detector errors, f Luorophore intensity, off-target binding by affinity reagents, and protein detection. It is normalized to correct for the causes of errors and biases, including gender.

[0013] In another context, in this specification, candidate proteins in a sample of an unknown protein are A computer-based method for identification is disclosed, and the method involves the following steps. (a) Multiple affinity reagent probes for the unknown protein in the sample The process involves receiving information on the coupling measurement for each of the following using the computer. Each affinity reagent probe is one or more candidate proteins among several candidate proteins. (b) A process configured to selectively bind to a coprotein; (b) Information on the binding measurement At least a portion of the report is compared against a database containing multiple protein sequences. A step of comparing using a pewter, wherein each protein sequence is compared with the plurality of candidate proteins A step corresponding to one candidate protein in the substance; and (c) prior to the information of the binding measurement. The above applies to the database containing at least a portion of the plurality of protein sequences. Based at least on comparison, one or more candidate proteins are selected from the plurality of candidate proteins. A process to remove plaque.

[0014] In some embodiments, the step of removing the one or more candidate proteins is performed before It is based at least on the predetermined criteria for the binding measurement related to the candidate protein. In one embodiment, the predetermined criterion is the first of the plurality of affinity reagent probes. The one or more candidate proteins having a binding measurement below a predetermined threshold for the number Includes. In some embodiments, the plurality of affinity reagent probes include 50 or fewer affinity reagents. Includes drug probes. In some embodiments, the plurality of affinity reagent probes are 100 The following affinity reagent probes are included. In some embodiments, the plurality of affinity reagent probes The probes contain 500 or fewer affinity reagent probes.

[0015] In some embodiments, the method involves extracting one or more unknown proteins from the sample. The process further includes the step of generating a paper or electronic report to identify. In some embodiments, The sample includes a biological sample. In some embodiments, the biological sample is , obtained from the target. In some embodiments, the method is the identified candidate protein The method further includes identifying the disease state in the subject, based at least on the above.

[0016] Any additional aspects and advantages of this disclosure are shown and described only in the manner illustrated in this disclosure. As will be readily apparent to those skilled in the art from the following detailed explanation, Disclosure may take other and different forms, and some of its details may vary. In all clear respects, this disclosure may be modified without any deviation whatsoever. Therefore, the drawings and descriptions should be considered, in essence, illustrative and limited to... It should not be considered as an object.

[0017] Inclusion by reference All publications, patents, and patent applications referred to herein are subject to the individual publications. Each patent or patent application is incorporated by reference specifically and individually. Incorporated herein by reference to the same extent as shown herein. Any publications and patents or patent applications incorporated by reference that contradict this disclosure are subject to the terms of this disclosure. With regard to scope, this specification takes precedence over any such conflicting matters, and / Alternatively, the intention is to be in a superior position. [Brief explanation of the drawing]

[0018] Novel features of the present invention are described in particular in the appended claims. Features of the present invention A better understanding of the advantages and other benefits can be found in the following section, which describes illustrative embodiments in which the principles of the present invention are utilized. For a detailed explanation, please refer to the following attached drawings (hereinafter also referred to as "Figures"). This can be obtained by referring to the "Figure (FIG.)" (also known as "Figures").

[0019] [Figure 1] The following are illustrative flowcharts illustrating the protein identification of unknown proteins in biological samples, following several embodiments. [Figure 2] This specification describes a computer-controlled system programmed to perform the methods provided herein, or, in other circumstances, configured to do so. [Figure 3] The performance of censored versus uncensored protein identification approaches, according to several embodiments, is demonstrated. [Figure 4] This paper demonstrates the acceptability of censored and uncensored protein identification approaches to randomized "false-negative" binding outcomes in several aspects. [Figure 5]This paper demonstrates the acceptability of censored and uncensored protein identification approaches to randomized "false-positive" binding outcomes in several aspects. [Figure 6] The performance of censored and uncensored protein identification approaches using overestimated or underestimated affinity reagent binding probabilities, according to several embodiments, is demonstrated. [Figure 7] The performance of censored and uncensored protein identification approaches using affinity reagents with unknown binding epitopes, according to several embodiments, is demonstrated. [Figure 8] The performance of censored and uncensored protein identification approaches using affinity reagents lacking binding epitopes, according to several embodiments, is demonstrated. [Figure 9] The performance of censored and uncensored protein identification approaches using affinity reagents targeting the top 300 most abundant trimers in the proteome, 300 randomly selected trimers in the proteome, or 300 least abundant trimers in the proteome, according to several embodiments, is demonstrated. [Figure 10] The performance of censored and uncensored protein identification approaches using affinity reagents having random off-target sites or biosimilar off-target sites, according to several embodiments, is demonstrated. [Figure 11] The performance of censored and uncensored protein identification approaches using optimal affinity reagent (probe) sets according to several embodiments is demonstrated. [Figure 12] The performance of censored and uncensored protein identification approaches using unmixed candidate affinity reagents and mixtures of candidate affinity reagents, according to several embodiments, is demonstrated. [Figure 13]Two hybridization steps for enhancing the binding between affinity reagents and proteins, according to several embodiments, are shown. [Modes for carrying out the invention]

[0020] Detailed explanation Various aspects of the present invention are shown and described herein, however such It will be obvious to those skilled in the art that the embodiments are provided merely as examples. Numerous changes and modifications may occur. Modifications and substitutions can be conceived by those skilled in the art without departing from the present invention. It is understood that various alternatives to the embodiments of the present invention described herein may be employed. It should.

[0021] The term "sample" as used herein generally refers to a biological sample (for example) For example, it refers to a sample containing protein. The sample is taken from tissue or cells, or from tissue or It may be collected from the cellular environment. In some cases, the sample is a tissue biopsy, blood, Plasma, extracellular fluid, dried blood spots, cultured cells, culture media, discarded tissue, plant matter synthetic proteins, bacterial and / or viral samples, fungal tissues, archaea, if The sample may contain or be derived from protozoa. Before collection, the sample should be examined by the source. It may be isolated from the subject. The sample may contain forensic evidence. A non-limiting example is the collection. Fingerprints, saliva, urine, blood, feces, semen, or other bodily fluids isolated from the primary source before use. This includes. In some cases, the protein is the primary source (cell) during sample preparation. They are isolated from bodily fluids such as tissues and blood, and from environmental samples. Samples are often derived from extinct species. , but not limited to, samples derived from fossils. Proteins are the primary source. It may be purified from, or it may not be purified, or otherwise, one The next source may or may not be concentrated. The primary source is homogenized before further processing. In some cases, Cells are lysed using a buffer such as RIPA buffer. Denaturation buffers are also used at this stage. It may be used. The sample is filtered or centrifuged to remove lipids and particulate matter. They may be separated. The sample may also be purified to remove nucleic acids, or RNases. It may also be treated with DNase. The sample may be untreated protein, denatured protein, or It may contain protein fragments or partially broken-down proteins.

[0022] Samples may be taken from subjects with a disease or disorder. The disease or disorder is infectious. Diseases, immune disorders or immune diseases, cancer, genetic disorders, degenerative diseases, lifestyle-related diseases, wounds, rare It may be a minor illness or an age-related illness. Infectious diseases are caused by bacteria, viruses, fungi, etc. It can be caused by a parasite. Non-limited examples of cancer include bladder cancer, lung cancer. Cancer, brain cancer, melanoma, breast cancer, non-Hodgkin lymphoma, cervical cancer, ovarian cancer, colon cancer, Rectal cancer, pancreatic cancer, esophageal cancer, prostate cancer, kidney cancer, skin cancer, leukemia, thyroid cancer This includes liver cancer and uterine cancer. Some examples of hereditary diseases or disorders are: While not limited to these, cystic fibrosis, Charcot-Marie-Tooth disease, and Huntington's disease are examples of conditions that can cause this. Ton's disease, Peutz-Jeghers syndrome, Down syndrome, rheumatoid arthritis, and Tay Sac syndrome. This includes lifestyle-related diseases. Non-specific examples of lifestyle-related diseases include obesity, diabetes, arteriosclerosis, heart disease, and stroke. Hypertension, cirrhosis, nephritis, cancer, chronic obstructive pulmonary disease (COPD), hearing problems, and chronic back pain. Including pain. Some examples of wounds, though not limited to, include abrasions, brain injuries, contusions, Burns, concussion, congestive heart failure, injuries at construction sites, dislocations, chest instability, fractures, hemothorax, intervertebral disc Herniated disc, hip pointer, hypothermia, laceration, pinched nerve e) pneumothorax, rib fractures, sciatica, spinal cord injury, tendon, ligament, and fascia injuries, traumatic brain injury, and whiplash. Samples should be taken before and / or after treatment of subjects with disease or disability. The sample may be collected later. The sample may be collected before and / or after the procedure. Samples may be taken during or between treatments. Multiple samples may be used to monitor the effects of the treatment over time. The sample may be taken from the subject for monitoring. The sample may not have diagnostic antibodies available. Samples may be collected from individuals who are known to have or suspected to have an infectious disease. .

[0023] The sample may be taken from a person suspected of having a disease or disorder. Symptoms such as fatigue, nausea, weight loss, ache and pain, weakness, or amnesia. Samples may be taken from subjects experiencing unexplained symptoms. The sample may be taken from a subject with an explainable symptom. Family medical history, age, environmental exposure, lifestyle risk factors, or other known risk factors From subjects at risk of developing diseases or disabilities caused by factors such as the presence of children, You may do so.

[0024] The sample may be taken from an embryo, fetus, or pregnant woman. In some cases, the sample is, It may contain proteins isolated from the mother's plasma. In some cases, the mother's blood Proteins isolated from circulating fetal cells in fluid.

[0025] Samples may be taken from healthy individuals. In some cases, the samples may be from the same individual. They may be collected over a long period of time. In some cases, samples obtained over a long period of time may be individual With the goal of monitoring physical health and detecting health problems early, analysis is performed. It may be done. In some embodiments, the sample may be collected at home or in a clinical setting. And then, prior to analysis, it may be transported by mail, courier, or other means of transport. For example, a user at home can collect a blood spot sample by finger prick. The blood spot sample is then dried and subsequently mailed before analysis. It is acceptable to transport them in this manner. In some cases, samples obtained over a long period of time are subject to healthy exercise. To monitor responses to stimuli that are expected to affect abilities or cognitive abilities, It may be used. Non-exclusive examples include responses to medicine, dietary therapy, or exercise therapy. include.

[0026] The sample protein was treated to remove any modifications that could interfere with epitope binding. For example, proteins are treated with glycosidase to remove post-translational glycosylation. It is acceptable to treat proteins with reducing agents to reduce the disulfide bonds within the protein. It is acceptable to treat it. Proteins are treated with phosphatase to remove phosphate groups. Other non-exclusive examples of post-translational modifications that can be removed include acetate, amide groups, and methyl groups. Lipids, ubiquitin, myristoylation, palmitoylation, isoprenylation or prenylation Modification (e.g., farnesol and geranylgeraniol), farnesylation, geranyl Geranylation, glypiation, lipoylation, flavin moiety attachment, ho This includes supanteteinylation and retynidenecif base formation. The sample also undergoes translation. Post-protein modification may be preserved by processing. In some cases, phosphatase Inhibitors may be added to the sample. In some examples, to protect the disulfide bond, An oxidizing agent may be added for this purpose.

[0027] The protein in the sample may be completely or partially denatured. In some embodiments, Proteins can be completely denatured by surfactants, strong acids or strong bases, and concentrated Reduced inorganic salts, organic solvents (e.g., alcohol or chloroform), irradiation, Alternatively, the protein may be denatured by the application of external stress such as heat. The protein is denatured with a denaturation buffer. The protein may be denatured by the addition of [a certain substance]. The protein may also be precipitated, freeze-dried, or [another substance] in a denaturation buffer. The protein may be suspended. The protein may be denatured by heating. Modification methods that are less likely to produce ornamentation may be selected.

[0028] The sample protein produces shorter polypeptides through conjugation. It may be processed either before or after. The remaining protein is used to generate fragments. It can be partially digested with enzymes such as rotinase K, or it can be left intact. In further examples, proteins are exposed to proteases such as trypsin. Examples of additional proteases include serine protease, cysteine ​​protease, and threoprotease. Nin protease, aspartate protease, glutamate protease, metallop It may contain rotease and asparagine peptidylase.

[0029] In some cases, extremely large and small proteins (e.g., titin) Removing such proteins can be useful, and such proteins can be removed by filtration or other appropriate methods. They can be removed. In some cases, extremely large proteins can be removed in amounts of 400 kilodull. n (kD), 450 kD, 500 kD, 600 kD, 650 kD, 700 kD, 750 kD, 800 kD, or 850 kD It may contain proteins exceeding a certain amount. In some cases, extremely large proteins are approximately 8,000 amino acids, approximately 8,500 amino acids, approximately 9,000 amino acids, approximately 9,500 amino acids, approximately 10,000 amino acids Acids, proteins with approximately 10,500 amino acids, approximately 11,000 amino acids, or more than approximately 15,000 amino acids. It may include. In some examples, small proteins are less than approximately 10 kD, less than 9 kD, and 8 kD. Less than kD, less than 7 kD, less than 6 kD, less than 5 kD, less than 4 kD, less than 3 kD, less than 2 kD, or 1 kD It may contain less than 50 units of protein. In some examples, small proteins are about 50 units. Less than 0 amino acids, less than 45 amino acids, less than 40 amino acids, less than 35 amino acids, or less than approximately 30 amino acids. It may contain proteins. Extremely large or small proteins are excluded by size exclusion chromatography. They can be removed by chromatography. Extremely large proteins can be removed by size exclusion chromatography. Isolated by Fee and treated with proteases to produce intermediate-sized polypeptides The sample may be processed and then recombined with an intermediate-sized protein.

[0030] The proteins in the sample are, for example, identifiable, in order to allow for sample duplication. Tags may be used for tagging. Some non-exclusive examples of identifiable tags include: Fluorophores, magnetic particles, or DNA barcoded base linkers are used. Fluorophores include GFP, YFP, RFP, eGFP, mCherry, tdtomato, FITC, and Alexa Fluor 350. , Alexa Fluor 405, Alexa Fluor 488, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluo r 555, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 647, Alexa Fluor 680, Alexa Fluor 750, Pacific Blue, Coumarin, BODIPY FL, Pacific Green, Oregon Green, Cy3 Cy5, Pacific Orange, TRITC, Texas Red, phycoerythrin, allophycocyanin (A Fluorescent proteins such as llophcocyanin, or other fluoroproteins known in the art. You may include a forehand.

[0031] Any number of protein samples can be duplicated. For example, the duplicated reaction can be 2, 3, 4 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, approximately 20, approximately 25, approximately 30, approximately 35 Approximately 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100 , or may include proteins from more than 100 initial samples. Identifiable tags are provided for each protein. It may provide a means to examine the substance with respect to the sample from which it originates, or to examine the substance from different samples. The protein may be induced to be isolated in different areas or solid supports. In one embodiment, the protein is then subjected to a mechanism to chemically bond the protein to the substrate. Applied to a modified substrate.

[0032] Any number of protein samples, without being tagged or duplicated. They may be mixed before analysis. For example, multiplexed reactions may involve 2, 3, 4, or 5 reactions. 6 pieces, 7 pieces, 8 pieces, 9 pieces, 10 pieces, 11 pieces, 12 pieces, 13 pieces, 14 pieces, 15 pieces, 16 pieces, 17 pieces, 18 pieces, 19 pieces, About 20 pieces, about 25 pieces, about 30 pieces, about 35 pieces, about 40 pieces, about 45 pieces, about 50 pieces, about 55 pieces, about 60 pieces, about 65 pieces, Initial samples of approximately 70, 75, 80, 85, 90, 95, 100, or more than 100. It may contain proteins. For example, diagnostics for rare conditions are pooled This may be performed on the sample. Analysis of individual samples may be performed after testing if the diagnosis is positive. This procedure may only be performed on samples from the pool. The samples are combinatorial pooled. The design may be duplicated without tagging, and in the design, individual The signal from the sample is analyzed from a pool using computer-based multiplexing. The samples are mixed into the pool in a manner that allows for differentiation.

[0033] The term "substrate," as used herein, generally refers to a solid support. It refers to a substrate that can do so. A substrate, or solid substrate, is a material in which proteins can also be covalently bonded. This can refer to any solid surface that can be non-covalently bonded. Non-limiting examples of solid substrates include grains. Children, beads, slides, surfaces of device components, films, flow cells, wells, chambers, Macrofluidic chamber, microfluidic chamber, channel, microfluidic channel, or includes any other surface. The surface of the substrate may be flat or curved, and It may have other shapes and may be smooth or uneven. The surface of the substrate is It may contain Icrowell. In some embodiments, the substrate is glass, dextran, etc. Any carbohydrate, plastics such as polystyrene or polypropylene, polyacrylic It may be composed of amide, latex, silicon, metals such as gold, or cellulose, and To enable or strengthen covalent or noncovalent bonds in proteins, further modification It may be decorated. For example, the surface of the substrate may be decorated with maleic acid moety or succinic acid moety. It may be functionalized by modification with specific functional groups such as amino groups, thiol groups, Alternatively, modification with chemical reactive groups such as acrylate groups, such as silanization, It may also be derivatized. Suitable silane reagents include aminopropyltrimethoxysilane, Contains minopropyltriethoxysilane and 4-aminobutyltriethoxysilane. The material may be functionalized with N-hydroxysuccinimide (NHS) functional groups. The glass surface is For example, epoxysilane, acrylatesilane, or acrylamidesilane can be used. It can be derivatized with other reactive groups such as acrylates or epoxys. Protein binding For this purpose, the substrate and processing preferably involve repeated bonding, washing, imaging, and dissolving. It is stable against the output process. In some examples, the substrate is a slide, flow cell, or microscale or nanoscale structures (e.g., microwells, micro Rules such as chloropillars, single-molecule arrays, nanoballs, nanopillars, or nanowires It can be the correct structure.

[0034] The spacing of functional groups on the substrate may be regular or random. Regular arrays are used in, for example, photolithography, dip-pen nanolithography, Nanoimprint lithography, nanosphere lithography, nanoball lithography, Nanopillar arrays, nanowire lithography, scanning probe lithography, thermochemical lithography Graphics, thermal scanning probe lithography, localized oxide nanolithography, molecular self-assembly, They may be prepared by stencil lithography or electron beam lithography. Regular Each functional group in the array is separated from any other functional group by 200 nanometers (nm). ) less than, or approximately 200 nm, approximately 225 nm, approximately 250 nm, approximately 275 nm, approximately 300 nm, approximately 32 5 nm, approx. 350 nm, approx. 375 nm, approx. 400 nm, approx. 425 nm, approx. 450 nm, approx. 475 nm, approx. 500 nm, approx. 52 5 nm, approx. 550 nm, approx. 575 nm, approx. 600 nm, approx. 625 nm, approx. 650 nm, approx. 675 nm, approx. 700 nm, approx. 72 5 nm, approx. 750 nm, approx. 775 nm, approx. 800 nm, approx. 825 nm, approx. 850 nm, approx. 875 nm, approx. 900 nm, approx. 92 5 nm, approx. 950 nm, approx. 975 nm, approx. 1000 nm, approx. 1025 nm, approx. 1050 nm, approx. 1075 nm, approx. 1100 nm , about 1125 nm, about 1150 nm, about 1175 nm, about 1200 nm, about 1225 nm, about 1250 nm, about 1275 nm, approx. 1300 nm, approx. 1325 nm, approx. 1350 nm, approx. 1375 nm, approx. 1400 nm, approx. 1425 nm, approx. 1450 nm, approx. 1 475 nm, approx. 1500 nm, approx. 1525 nm, approx. 1550 nm, approx. 1575 nm, approx. 1600 nm, approx. 1625 nm, approx. 1650 nm, approx. 1675 nm, approx. 1700 nm, approx. 1725 nm, approx. 1750 nm, approx. 1775 nm, approx. 1800 nm, approx. 1825 nm , about 1850 nm, about 1875 nm, about 1900 nm, about 1925 nm, about 1950 nm, about 1975 nm, about 2000 nm, Alternatively, they may be arranged so that they are greater than 2000 nm. Functional groups at random intervals are functional groups , on average at least about 50 nm, about 100 nm, about 150 nm, about 200 nm from any other functional group, approx. 250 nm, approx. 300 nm, approx. 350 nm, approx. 400 nm, approx. 450 nm, approx. 500 nm, approx. 550 nm, approx. 600 nm, approx. 650 nm, approx. 700 nm, approx. 750 nm, approx. 800 nm, approx. 850 nm, approx. 900 nm, approx. 950 nm, approx. 1000 nm They may be provided in a densely packed state such as being greater than 100 nm.

[0035] The substrate may be indirectly functionalized. For example, the substrate may be PEGylated and functionalized This may be applied to all or a subset of PEG molecules. The substrate is micros Kale or nanoscale structures (e.g., microwells, micropillars, single-molecule structures) Due to the regular structure of subarrays, nanoballs, nanopillars, or nanowires. It may be functionalized using technology suitable for the purpose.

[0036] The substrate includes metal, glass, plastic, ceramic, or a combination thereof. , may include any material. In some preferred embodiments, the solid substrate is in a flow cell It's possible. A flow cell can consist of a single layer or multiple layers. For example, flow The cell consists of a base layer (for example, one made of borosilicate glass), and a channel layer covering the base layer. This may include, for example, etched silicon, and a cover layer or top layer. When those layers are assembled together, an enclosed channel may be formed, and the channel is It has an inlet / outlet at both ends through the cover. The thickness of each layer is variable, but preferred The thickness is less than approximately 1700 μm. The layer is not limited to, but may include photosensitive glass or bronzes. This technology includes acid glass, fused silicate, PDMS, or silicon. It can be composed of any suitable material known in the context. Different layers may be made from the same material, or It can be composed of different materials.

[0037] In some embodiments, the flow cell has an opening for a channel at the bottom of the flow cell. It may include parts. The flow cell has millions of attached parts at positions that can be visualized separately. It may include a target conjugation site. In some embodiments, it may be used in the present invention. The various flow cells used have different numbers of channels (e.g., 1 channel, 2 or more). Channels of 0, 3 or more channels, 4 or more channels, 6 or more channels, 8 or more channels Channels, 10 or more channels, 12 or more channels, 16 or more channels, or more than 16 The channels may include various flow cells with channels of different depths or widths. These may include, and these may differ between channels in a single flow cell, or may be different. The channels of the flow cell may differ. One channel may also be different in depth. and / or width may change. For example, one channel may be one or At multiple locations, at depths of less than approximately 50 μm, approximately 50 μm, less than approximately 100 μm, and approximately 100 μm Depths of approximately 100 μm to approximately 500 μm, approximately 500 μm, or greater than approximately 500 μm It could be. The channel is not limited to, but can be circular, semicircular, rectangular, or base. It may have any cross-sectional shape, including a triangular or oval cross-section.

[0038] Proteins may be spotted onto a substrate, dropped onto a substrate, or pipetteted. It may be poured, washed, or otherwise applied. For example, NHS esters. In the case of substrates functionalized with Moeti, protein modification is not required. Substrates functionalized with moety (e.g., sulfhydryl, amine, or linker DNA) In that case, a crosslinking reagent (e.g., disuccinimidyl suberate, NHS, sulfonamide) is used. ) may be used. In the case of a substrate functionalized with linker DNA, the sample protein is They may be modified with complementary DNA tags. In some cases, the protein is in the electrostatic phase. It may be functionalized so that it can bond to the substrate through interaction.

[0039] Photoactivating crosslinking agents are used to direct the crosslinking of a sample to a specific area on the substrate. i. The photoactivating crosslinking agent attaches each sample to a known area of ​​the substrate, thereby transferring the protein It may be used to enable sample duplication. Photoactivating crosslinking agents include, for example, proteins. By detecting the fluorescent tag before the protein crosslinks, successfully tagged proteins can be identified. This can enable the specific adhesion of the material. Examples of photoactivatable crosslinking agents include, but are not limited to, N- 5-Azido-2-nitrobenzoyloxysuccinimide, sulfosuccinimidyl 6-(4'-A (Zido-2'-nitrophenylamino)hexanoate, succinimidyl 4,4'-adipentanoate Sulfosuccinimidyl 4,4'-adipentanoate, succinimidyl 6-( Dipentanamide) hexanoate, sulfosuccinimidyl 6-(4,4'-adipentanamide) (d) Hexanoate, succinimidyl 2-((4,4'-adipentanamide)ethyl)-1,3'-diethyl Opropionate and sulfosuccinimidyl 2-((4,4'-adipentanamide)ethyl Contains )-1,3'-dithiopropionate.

[0040] Polypeptides may be attached to a substrate by one or more residues. Some examples In this process, polypeptides are synthesized via the N-terminus, C-terminus, both ends, or through internal residues. And it may adhere to it.

[0041] In addition to permanent crosslinking agents, photocleavable linkers are used for some applications. And by doing so, it becomes possible to selectively extract proteins from the substrate after analysis. It may be appropriate to do so. In some cases, photoclear crosslinking agents are several It may be used for different multiplexed samples. In some cases, the photoclear crosslinking agent is multi It may be used from one or more samples in the combined reaction. In some cases The multiplexed reaction involved crosslinking the substrate to a control sample via a permanent crosslinking agent, and then cutting the light. The experimental sample may include a substrate crosslinked via a deinterrupting crosslinking agent.

[0042] Each conjugated protein is optically connected to the other conjugated protein. Spatially separate each of the other conjugated proteins so that they can be resolved. It is permissible to label proteins individually using their unique spatial addresses. This may be done. In some embodiments, this means that each protein molecule is connected to other protein molecules. Low concentrations of protein and low density on the substrate are used to spatially separate them. This can be achieved by conjugation using an attachment site. For example, photoactivated crosslinking agents. When used, light is used so that the protein is attached to a predetermined position. Patterns may be used.

[0043] In some embodiments, each protein is associated with a unique spatial address. For example, when proteins attach to a substrate at positions where they are spatially separated, each tan A pak can be assigned an indexed address based on coordinates, etc. In this example, the grid of pre-assigned unique spatial addresses is predetermined It may be fixed. In some embodiments, the arrangement of each protein is fixed to a mark on the substrate. The substrate may include easily identifiable fixed marks so that it can be determined against. In the example, the substrate has grid lines and / or and permanently marked on its surface. There may be a “starting point” or other criteria. In some examples, the surface of the substrate is crosslinked. To provide a reference for identifying the location of proteins, permanently or semi-permanently It is acceptable for it to be marked. Patterns such as the outer edge of a conjugated polypeptide. Its own shape is also used as a criterion for determining the unique position of each spot. good.

[0044] The substrate may also include conjugated protein standards and controls. The conjugated protein standards and controls were conjugated at known positions. It may be a peptide or protein of a sequence of knowledge. In some examples, conjugate Prepared protein standards and controls can serve as internal controls in assays. The protein may be applied to the substrate from a purified protein stock, or Nucleic Acid-Programmable Protein Array (N) It may be synthesized on a substrate by processing such as APPA.

[0045] In some examples, the substrate may include fluorescent standards. These fluorescent standards are These fluorescent standards may be used to calibrate the intensity of the fluorescence signal between samples. Furthermore, in order to correlate the intensity of the fluorescence signal with the number of fluorophores present in a certain region... , may be used. Fluorescent standards are different types of fluoresceins used in assays. It may include some or all of the fores.

[0046] Once the substrate is conjugated with the protein from the sample, multiple affinity reagent measurements are performed. The measurement process described herein can be performed on various affinity tests. Drugs may be used. In some embodiments, multiple affinity reagents may be mixed together. Furthermore, the measurement involves the binding of the affinity reagent mixture to the protein-substrate conjugate. And it is permissible to implement it.

[0047] The term "affinity reagent" as used herein generally refers to a protein... This refers to a reagent that binds to a peptide with reproducible specificity. For example, affinity reagents. The drug was an antibody, antibody fragment, aptamer, miniprotein binder, or peptide. It is acceptable. In some embodiments, the mini protein binder has a length of 30 to 210 amino acids. It may contain a protein binder which may be between acids. In some embodiments, An protein binder may be designed. In some embodiments, a monoclonal antibody It can be selected. In some cases, antibody fragments such as Fab fragments can be selected. In some cases, the affinity reagent may be a commercially available affinity reagent such as a commercially available antibody. In that case, the desired affinity reagent is used to identify those with useful characteristics. The affinity reagents may be selected by screening for commercial use.

[0048] Affinity reagents may have high, intermediate, or low specificity. In some examples... Affinity reagents may recognize several different epitopes. In some examples, Affinity reagents may recognize epitopes present in two or more different proteins. How many? In that example, affinity reagents recognize epitopes present in many different proteins. It may be done. In some cases, the affinity reagent used in the method of this disclosure is epi It may be highly specific to a single tope. In some cases, this disclosure The affinity reagent used in this method is for only one epitope containing post-translational modifications. , may be highly specific. In some cases, affinity reagents may be highly similar. It may have specificity to the epitope. In some cases, highly similar epitopes Affinity reagents that have specificity for a protein candidate sequence are highly similar to a 1A Designed specifically to distinguish between candidates (or isoforms) with minoic acid mutations. Obtain. In some cases, affinity reagents maximize the coverage of the protein sequence. Therefore, it may have specificity to highly diverse epitopes. In some embodiments Furthermore, due to the probabilistic nature of probe binding to protein substrates, the results may vary. It is expected that this may provide additional information regarding protein identification. Therefore, the experiment may be repeated using the same affinity probe.

[0049] In some cases, a specific single epitope is recognized by affinity reagents. Alternatively, multiple epitopes do not need to be completely known. For example, one affinity reagent Or specific to multiple full-length proteins, protein complexes, or protein fragments. Regarding binding, it may be designed or selected without knowledge of specific binding epitopes. Qualitative The process may have resulted in a more sophisticated binding profile for this reagent. Even if the specific binding epitope is unknown, the binding measurement using the affinity reagent is It can be used to determine the identity of a protein. For example, protein labels A commercially available antibody or aptamer designed to bind precisely is used as the affinity reagent. This is possible under assay conditions (e.g., fully folded, partially denatured, After qualitative analysis under conditions (which are completely denatured), this affinity reagent for the unknown protein The binding may provide information about the identity of the unknown protein. In this case, a group of protein-specific affinity reagents (e.g., commercially available antibodies or apters) Mar) along with knowledge about the specific epitopes they target, or the knowledge It can be used to generate protein identification, either by deeming it or not. In some cases In this case, the group of protein-specific affinity reagents was 50, 100, 200, 300, 400, 500 pieces, 600 pieces, 700 pieces, 800 pieces, 900 pieces, 1000 pieces, 2000 pieces, 3000 pieces, 4000 pieces, 5000 pieces, 10000 pieces It may contain 1, 20,000, or more than 20,000 affinity reagents. In some cases, The affinity reagents are a group of all cities in which reactivity to targets has been proven in specific organisms. It may include commercially available affinity reagents. For example, a group of protein-specific affinity reagents may include each affinity The assay may be performed sequentially using binding measurements that are individually performed for each sex reagent. In that case, a subset of protein-specific affinity reagents is mixed before the binding measurement. For example, for each binding measurement path, a novel mixture of affinity reagents may be used. Selected to include a subset of affinity reagents randomly chosen from the entire set. For example, each of the following mixtures contains many affinity reagents present in multiple mixtures. It is expected that they will be generated in the same random manner. In some cases, Protein identification is performed more rapidly using a mixture of protein-specific affinity reagents. Obtain. In some cases, such mixtures of protein-specific affinity reagents are available. The percentage of unknown proteins to which affinity reagents bind in each individual pathway is determined. It can be increased. The affinity reagent mixture is 1%, 5%, 10% of all available affinity reagents. It may contain 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or more than 90% in a single experiment. The mixture of affinity reagents to be evaluated may have common affinity reagents, or They don't have to be common. In some cases, within a group that binds to the same protein, multiple A number of affinity reagents of different types may exist. In some cases, each affinity reagent in the population It may bind to different proteins. Multiple affinity proteins that have affinity for the same protein. When a reagent binds to a single unknown protein, the common target of the affinity reagent is unknown. The trustworthiness of intelligence proteins in identity can increase. In some cases... Therefore, using multiple protein affinity reagents that target the same protein is effective for multiple affinity The reagent binds to different epitopes on the same protein and targets the protein. Only a subset of affinity reagents bind to the binding epitope, which is a post-translational modification or other stereochemistry of the binding epitope. It can provide redundancy in cases where it could be interrupted by a steric hinderance. In that case, the binding of an affinity reagent whose binding epitope is unknown is due to the binding epitope By combining this with binding measurements of affinity reagents with known tope properties, protein identification can be generated. It can be used for that purpose.

[0050] In some cases, one or more affinity reagents are used in quantities of 2, 3, 4, 5, 6, or 7. It binds to amino acid motifs of a predetermined length, such as 1, 8, 9, 10, or more than 10 amino acids. They may be selected to match. In some examples, one or more affinity reagents may be 2 To bind to amino acid motifs of varying lengths, from 1 to 40 amino acids. You deserve to be chosen.

[0051] In some cases, affinity reagents may be labeled with DNA barcodes. In this example, the DNA barcode may be used to purify the affinity reagent after use. In some cases, DNA barcodes sort affinity reagents for repeated use. It may be used for that purpose. In some cases, affinity reagents may be used after affinity testing. The reagents may be labeled with fluorophores that can be used to sort them.

[0052] A family of affinity reagents may include one or more types of affinity reagents. The methods of this disclosure include antibodies, antibody fragments, Fab fragments, aptamers, peptides, and proteins. A family of affinity reagents containing one or more of these substances may be used.

[0053] Affinity reagents may be modified. Modifications are not limited to those that affect detection moiety. This includes combination. The detection moety may be combined directly or indirectly. For example, detection moety The tee may be directly covalently bonded to the affinity reagent, or bonded via a linker. Often, or complementary, DNA tags or affinity pairs such as biotin-streptavidin. The binding may occur via a reaction. The binding may withstand light washing and elution of affinity reagents. A suitable method can be chosen.

[0054] Affinity reagents are used, for example, to identify or quantify binding events (e.g., the biometrics of binding events). It may be tagged with an identifiable tag that enables (by light detection). Some non-limiting examples include: fluorophores, fluorescent nanoparticles, quantum dots, Linkers based on magnetic nanoparticles or DNA barcodes. The fluorophores used are G FP, YFP, RFP, eGFP, mCherry, tdtomato, FITC, Alexa Fluor 350, Alexa Fluor 405, A Alexa Fluor 488, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 5 68, Alexa Fluor 594, Alexa Fluor 647, Alexa Fluor 680, Alexa Fluor 750, Pacific Blue, Coumarin, BODIPY FL, Pacific Green, Oregon Green, Cy3, Cy5, Pacific Orange TRITC, Texas Red, phycoerythrin, allophycocyanin, or in the art It may contain other known fluorescent proteins, such as fluorophores. Alternatively, affinity testing may be performed. Drugs are detected when their binding event is directly detected, for example, by SPR detection of the binding event. The tag may be removed.

[0055] The detected molecules are not limited to fluorophores, bioluminescent proteins, and DNA segments including mutation regions and barcode regions, or nanoparticles such as magnetic particles It may include a chemical tether for coupling. The detection mechanism is a different type of excitation or emission. It may contain several different fluorophores that have turns.

[0056] The detection moiety may be cleavable from the affinity reagent. This is no longer of interest. The process of removing the detection moiety from the affinity reagent reduces signal contamination. This can make that possible.

[0057] In some cases, affinity reagents are unmodified. For example, if affinity reagents are anti If it is a body, the presence of antibodies can be detected by atomic force microscopy. Affinity reagents are not yet available. The antibody may be modified and, for example, specific to one or more affinity reagents. It may be detected by obtaining it. For example, if the affinity reagent is a mouse antibody If present, mouse antibodies may be detected using anti-mouse secondary antibodies. Alternating between, affinity The reagent may be an aptamer detected by an antibody specific to the aptamer. The body may be modified with detection moety as described above. In some cases, secondary antibody The presence of [the substance] may be detected by atomic force microscopy.

[0058] In some cases, affinity reagents are the same modified, for example, conjugated green It may contain a fluorescent protein, or it may contain two or more different modifications. For example, Each affinity reagent has a different excitation wavelength or emission wavelength, and is used by several different fireflies. It may be conjugated to one of the photomoety. Several different affinity reagents can be combined. Since this can be done and / or identified, it may allow for the redundancy of affinity reagents. In one example, the first affinity reagent may be conjugated to a green fluorescent protein. The second affinity reagent may be conjugated with a yellow fluorescent protein, and the third affinity reagent The drug may be conjugated with a red fluorescent protein, and therefore these three affinity tests Drugs can be multiplyed and identified by their fluorescence. In further examples, the first The fourth and seventh affinity reagents may be conjugated to a green fluorescent protein, and the second The fifth and eighth affinity reagents may be conjugated to a yellow fluorescent protein, and The third, sixth, and ninth affinity reagents may be conjugated to a red fluorescent protein; In this case, the first, second, and third affinity reagents can all be duplicated, while the second and fourth, and the seventh, as well as the third, sixth, and ninth affinity reagents, two further multiplexed reactions The number of affinity reagents that can be multiplexed together is used to distinguish them. It can vary depending on the detection method. For example, affinity reagents labeled with fluorophores Multiplexing may be limited by the number of unique fluorophores available. Further examples include... Therefore, the multiplexing of affinity reagents labeled with DNA tags is determined by the length of the DNA barcode. You may do so.

[0059] The specificity of each affinity reagent can be determined before use in the assay. Maturity can be determined in control experiments using known proteins. The test method may be used to determine the specificity of the affinity reagent. In one example, the substrate may be treated with To evaluate the specificity of multiple affinity reagents by placing a known protein standard at a known position. It may be used for this purpose. In another example, the specificity of each affinity reagent is used for binding to the control and standard. The substrate can be calculated from and subsequently used to identify the experimental sample, so that it can be used for experimental sample identification. It may include both the material and the control and standard panels. In some cases, A known affinity reagent may be included together with an affinity reagent of known specificity. Data from affinity reagents for specificity of knowledge can often be used to identify proteins. Furthermore, the binding patterns of affinity reagents with unknown specificity to the identified proteins are It may be used to determine the binding specificity of each protein. To evaluate whether they matched, known binding data for other affinity reagents was used for any individual It is also possible to reconfirm the specificity of the affinity reagent. In some cases, The frequency of binding of affinity reagents to each known protein conjugated on the substrate is as follows: It can be used to derive the probability of binding to any of the proteins above. In some cases, known tangents containing epitopes (e.g., amino acid sequences or post-translational modifications) The frequency of binding to the protein determines the probability of affinity reagents binding to a specific epitope. It can be used for this purpose. Therefore, by using multiple affinity reagent panels, the characteristics of affinity reagents can be determined. The opposite sex can become increasingly refined with each iteration. Uniquely specific to a particular protein. While affinity reagents may be used, the methods described herein do not require them. It is not necessary. In addition, the method may be effective in a certain range of specificities. In example, the method described herein involves the affinity reagent being any particular protein It is not specific to the substance, but instead uses amino acid motifs (e.g., tripeptides). This may be particularly effective when it is specific to AAA.

[0060] In some cases, affinity reagents have high, intermediate, or low binding affinity. They may be selected in such a way. In some cases, a parent with low or intermediate binding affinity may be selected. A compatibility reagent may be selected. In some cases, the affinity reagent is about 10 -3 M, 10 -4 M, 10 -5 M, 10 -6 M, 10 -7 M, 10 -8 M, 10 -9 M, 10 -10 of M or 10 -10 a dissociation constant smaller than M may have. In some cases, the affinity reagent is about 10 -10 M, 10 -9 M, 10 -8 M, 1 0 -7 M, 10 -6 M, 10 -5 M, 10 -4 M, 10 -3 M, 10 -2 exceeding M or 10 -2 a dissociation constant larger than M may have. In some cases, a low or intermediate k off rate or an intermediate or high k on rate affinity reagent may be preferred.

[0061] Some of the affinity reagents may be selected to bind to a modified amino acid sequence such as a phosphorylated or ubiquitinated amino acid sequence . In some examples, one or more affinity reagents may be selected to be broadly specific for a family of epitopes that may be included by one or more proteins . In some examples, one or more affinity reagents may bind to two or more different proteins . In some examples, one or more affinity reagents may bind weakly to one or more of their targets . For example, the affinity reagent may bind to one or more of its targets at less than 10%, less than 10%, less than 15%, less than 20%, less than 25%, less than 30%, or less than 35% . In some examples, one or more affinity reagents may bind moderately to one or more of their targets . For example, the affinity reagent may bind to one or more of its targets at less than 10%, less than 10%, less than 15%, less than 20%, less than 25%, less than 30%, or less than 35% . In some examples, one or more affinity reagents may bind moderately to one or more of their targets . It may be bound to or strongly bound. For example, affinity reagents may have concentrations of over 35%, over 40%, over 45%, over 60%, and 65%. More than %, More than 70%, More than 75%, More than 80%, More than 85%, More than 90%, More than 91%, More than 92%, More than 93%, More than 94%, More than 95%, More than 96% More than 97%, more than 98%, or more than 99% can bind to one or more of those targets.

[0062] To compensate for weak bonding, an excess of affinity reagent may be applied to the substrate. The affinity reagent is Approximately 1:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1, 9:1, or 10:1 ratio relative to the sample protein. It is acceptable to apply it in excess. The affinity reagents should be in ratios of approximately 1:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1. A ratio of 9:1 or 10:1 is used to overestimate the expected occurrence rate of epitopes in the sample protein. It may be applied to excesses.

[0063] To compensate for the rapid dissociation rate of affinity reagents, a linker moiety is bound to each affinity reagent. The linker moiety then transfers the bound affinity reagent to the substrate to which it binds. Alternatively, it may be used to reversibly ligate to an unknown protein, for example, a DNA tag. However, each affinity reagent can be bound to the end, and different DNA tags can be attached to the substrate or each unknown tag. It can bind to proteins. After the affinity reagent hybridizes to the unknown protein, The marker DNA, which is complementary at one end to the DNA tag bound to the affinity reagent, Furthermore, it is complementary at the other end to the tag that is bonded to the base material, but This can be added to the tip to bind affinity reagents to the substrate, and this is affinity This prevents the reagent from dissociating before measurement. After binding, the linked affinity reagent acts as a DNA linker. The bonds can be separated by washing with heat or in the presence of a high salt concentration to break them down.

[0064] Figure 13 shows several embodiments of enhancing the binding between affinity reagents and proteins. The two hybridization steps involved are shown. In particular, step 1 in Figure 13 is the process of using affinity reagents. Hybridization is observed. As shown in step 1, affinity reagent 1310 is used to convert proteins. Hybridize with protein 1330. Protein 1330 is bound to slide 1305. Process As shown in 1, affinity reagent 1310 has a bound DNA tag 1320. In this embodiment, the affinity reagent may have multiple bound DNA tags. In that embodiment, the affinity reagents are 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 1, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more than 20 It may have a bound DNA tag. DNA tag 1320 is an ssDNA tag having recognition sequence 1325. It includes. In addition, protein 1330 has two DNA tags 1340. In some embodiments, The DNA tag may be added using a chemical reaction that involves a reaction with cysteine ​​in the protein. In some embodiments, the protein has multiple bound DNA tags. i. In some embodiments, the protein is 1, 2, 3, 4, 5, 6, 7, 8 , 9 pieces, 10 pieces, 11 pieces, 12 pieces, 13 pieces, 14 pieces, 15 pieces, 16 pieces, 17 pieces, 18 pieces, 19 pieces, 20 pieces, 25 pieces, 3 0 pieces, 35 pieces, 40 pieces, 45 pieces, 50 pieces, 55 pieces, 60 pieces, 65 pieces, 70 pieces, 75 pieces, 80 pieces, 85 pieces, 90 pieces, 95 pieces It may have 1, 100, or more than 100 bound DNA tags. Each DNA tag 1340 is It contains an ssDNA tag having recognition sequence 1345.

[0065] As shown in step 2, the DNA linker 1350 interacts with affinity reagent 1310 and protein 1330. The DNA links 1320 and 1340, respectively, are hybridized. 50 contains ssDNA having sequences complementary to recognition sequences 1325 and 1345, respectively. As shown in step 2, recognition sequences 1325 and 1345 are linked to DNA tag 1320 by DNA linker 1350. To enable simultaneous binding of both and 1340, the DNA linker 1350 It is located. In particular, the first region 1352 of DNA linker 1350 is selectively hive with recognition sequence 1325. The DNA linker 1350 is reduced, and the second region 1354 of the DNA linker 1350 is selectively hive with the recognition sequence 1345. Reduce. In some embodiments, the first region 1352 and the second region 1354 are DN They may be positioned far apart from each other on the A linker. In particular, in some embodiments, DNA phosphor The first region of the Kerr and the second region of the DNA linker are located between the first and second regions. They may be positioned apart by a non-hybridized spacer array. Furthermore, In some embodiments, the recognition sequence is not entirely complementary to the DNA linker. It is possible, but still possible, to bind to the DNA linker sequence. In this case, the length of the recognition sequence is less than 5 nucleotides, 5 nucleotides, 6 nucleotides, 7 nucleotides Cleotide, 8 nucleotides, 9 nucleotides, 10 nucleotides, 11 nucleotides, 12 nucleotides Rheotide, 13 nucleotides, 14 nucleotides, 15 nucleotides, 16 nucleotides, 17 nucleotides Rheotide, 18 nucleotides, 19 nucleotides, 20 nucleotides, 21 nucleotides, 22 nucleotides Rheotide, 23 nucleotides, 24 nucleotides, 25 nucleotides, 26 nucleotides, 27 nucleotides Rheotide, 28 nucleotides, 29 nucleotides, 30 nucleotides, or more than 30 nucleotides This may be the case. In some embodiments, the recognition sequence is 1 for a complementary DNA tag sequence. It may have one or more inconsistencies. In some embodiments, the recognition sequence is 10 nucleos Approximately one of the cidos may be inconsistent with the complementary DNA tag sequence, but still... This can then hybridize to a complementary DNA tag sequence. In some embodiments, the recognition sequence This means that less than one out of 10 nucleotides may be inconsistent with the complementary DNA tag sequence, However, it can still hybridize to complementary DNA tag sequences. In some embodiments, The recognition sequence has approximately 2 out of 10 nucleotides that are mismatched with the complementary DNA tag sequence. It is possible, but still possible to hybridize to complementary DNA tag sequences. In that embodiment, the recognition sequence has more than two nucleotides out of 10 that correspond to the complementary DNA tag sequence. Although inconsistencies may occur, hybridization to complementary DNA tag sequences is still possible. ru.

[0066] The affinity reagent may also contain magnetic components. Manipulate several or all bound affinity reagents in the same image plane or z-stack. Therefore, it may be useful to manipulate some or all affinity reagents on the same image plane. This can improve the quality of image data and reduce noise in the system.

[0067] The term "detector," as used herein, generally refers to a device that detects a signal. This refers to a device that can detect the binding event of an affinity reagent to a protein. Includes a signal indicating presence or absence. The signal is a surface plasmon resonance (SPR) signal. This may be a direct signal indicating the presence or absence of a binding event, such as a null value. Signals are indirect signals that indicate the presence or absence of a binding event, such as fluorescent signals. It may be a signal. In some cases, the detector can detect a signal. , may include optical and / or electronic components. The term "detector" means detection method It may be used in law. Non-limiting examples of detection methods include optical detection, spectroscopic detection, and electrostatic detection. We use methods such as biomechanical detection, electrochemical detection, magnetic detection, fluorescence detection, and surface plasmon resonance (SPR). Includes, but is not limited to, fluorescence measurement and ultraviolet-visible absorbance. Optical detection methods include fluorescence measurement and ultraviolet-visible absorbance. This includes, but is not limited to, mass spectrometry and nuclear magnetic resonance (NMR). ) Including spectroscopy and infrared spectroscopy. Electrostatic detection methods are not limited to, This includes gel-based techniques, such as gel electrophoresis. Electrochemical detection methods are limited. Although it is not done, the electrons of the amplified product after high-performance liquid chromatography separation of the amplified product Includes atmospheric chemical detection.

[0068] Protein identification in the sample Proteins are essential components of the cells and tissues of living organisms. These substances typically produce a large set of various proteins, known as the proteome. The proteome can change over time and reflects the various experiences a cell or organism has. It can also change as a function of a tag (e.g., a stage of the cell cycle or a disease state). Large-scale studies of the proteome (e.g., experimental analysis) are referred to as proteomics and can be obtained. In proteomics, there are multiple methods for identifying proteins, including immunoassays (e.g., enzyme-linked immunosorbent assay (ELISA) and Western blot), mass spectrometry-based methods (e.g., matrix-assisted laser desorption ionization method (MALDI) and electrospray ionization method (ESI)), hybrid methods (e.g., mass spectrometry immunoassay (MSIA)), and protein microarrays. For example, single molecule proteomics techniques can be attempted to infer the identity of protein molecules in a sample by a variety of approaches ranging from direct functionalization of amino acids to the use of affinity reagents. Information or measurements collected from such approaches are typically analyzed by a suitable algorithm to identify the proteins present in the sample.

[0069] Accurate quantification of proteins can also face challenges due to lack of sensitivity, lack of specificity, and detector noise. In particular, accurate quantification of proteins in a sample can face challenges due to random and unpredictable variations in the detector signal level that can cause errors in identifying and quantifying proteins. In some cases, instruments and detection systems can be calibrated and errors removed by monitoring the diagnosis of the instrument and its behavior in common modes. However, protein binding (e.g., by affinity reagent probes) inherently has less than ideal sensitivity and binding specificity​ It is a probabilistic process.

[0070] This disclosure provides methods and systems for accurate and efficient identification of proteins. The methods and systems provided herein identify proteins in a sample. Errors during the process can be significantly reduced or eliminated. Such methods and systems can significantly reduce or eliminate errors. This enables accurate and efficient identification of candidate proteins in samples of unknown proteins. Protein identification is set to selectively bind to one or more candidate proteins. This can be obtained based on repeated calculations using information from the binding measurement of affinity reagent probes. Protein identification can be optimized so that it can be computed with a minimal memory footprint. Protein identification involves the presence of one or more candidate proteins in the sample. This may include a process for generating a reference level.

[0071] In one aspect, as described herein, candidate proteins in a sample of an unknown protein. A computer-based method 100 for repeatedly identifying is disclosed. (For example, as shown in Figure 1). The method involves multiple parental relationships for an unknown protein in the sample. The computer receives information on the binding measurement for each of the wattability reagent probes. The process may include steps such as step 105. In some embodiments, multiple affinity reagents The probe may include a pool of multiple individual affinity reagent probes. For example, affinity The pool of reagent probes includes 2 types, 3 types, 4 types, 5 types, 6 types, and 7 types. It may contain 8, 9, 10, or more than 10 affinity reagent probes. In one embodiment, the affinity reagent probe pool includes two types of affinity reagent probes. And so, the combination is the composition of affinity reagent probes in the pool of affinity reagent probes It accounts for the majority. In some embodiments, the affinity reagent probe pool consists of three types of affinity The combination may include affinity reagent probes, and the affinity of affinity reagent probes in the pool It constitutes the majority of the composition of the affinity reagent probe. In some embodiments, affinity reagent probe The pool may contain four types of affinity reagent probes, and the combination is such that the affinity reagent probe It constitutes the majority of the affinity reagent probe composition in the probe pool. In some embodiments, The affinity reagent probe pool may contain five types of affinity reagent probes, and the combination thereof This accounts for the majority of the composition of affinity reagent probes in the pool of affinity reagent probes. In one embodiment, the affinity reagent probe pool includes more than five types of affinity reagent probes. This combination may include a set of affinity reagent probes in a pool of affinity reagent probes. It accounts for the majority of the composition. Each affinity reagent probe is one of several candidate proteins. Alternatively, it can be configured to selectively bind to multiple candidate proteins. Affinity reagent profile The b can be an affinity reagent probe for the k-mer. In some embodiments, the affinity of the k-mer Each reagent probe contains one or more candidate proteins from a group of candidate proteins. It is configured to selectively bind to an unknown protein. The information from the binding measurement is used to determine which protein is bound to the unknown protein. This may include a set of probes that are considered to be such.

[0072] Next, at least a portion of the binding measurement information is in a database containing multiple protein sequences. The protein can be compared by computer (for example, step 110). Each column may correspond to one of a plurality of candidate proteins. The plurality of candidate proteins may include at least 10, at least 20, at least 30, at least 40 proteins, at least 50 proteins, at least 60 proteins, at least 70 proteins, at least 80 proteins, at least 90 proteins, at least 100 proteins, at least 150 proteins, at least 200 proteins, at least 250 proteins, at least 300 proteins, at least 350 proteins, at least 400 proteins, at least 450 proteins, at least 500 proteins, at least 600 proteins, at least 700 proteins, at least 800 proteins, at least 900 proteins, at least 1000 proteins, or more than

[0073] Next, for each one or more of the plurality of candidate proteins, the probability that the candidate protein is present in the sample can be calculated or generated by a computer (e.g., step 115). The calculation or generation can be performed repeatedly. Alternatively, the calculation or generation can be performed without repetition. The probability can be repeatedly generated based on a comparison to a database containing a plurality of protein sequences and information on the binding measurement of the candidate protein. Thus, the input to the algorithm can include a database of protein sequences and a set of probes that are thought to bind to the unknown protein. The output of the algorithm can include the probability that each

[0074] protein in the database can be present in the sample. In some embodiments, the probability calculated and output in step 115 can be expressed as follows: P(protein_i | probe[1, 2, …, n], This value represents the set of probes that bind to protein_i [1, 2, …, n], and protein_ Given the length of i (for example, the number of peptides), a given protein (protein_i) This indicates the probability that ) is present in the sample.

[0075] In some embodiments, the step of calculating the probability of outputting one or more affinity This may include calculating the product of the probabilities of a reagent (probe) landing on a protein. For example, If it has been detected that n probes bind to a protein, then different probes... The probability of each landing on a protein is given by P_landing_probe_1, P_landing_probe_2, …, P_ It can be expressed as landing_probe_n. Therefore, one or more affinity reagents (probe The product of the probabilities of each probe landing on a protein is Product(P_landing_probe_1, P_landing_probe) It can be expressed as (B_2, ..., P_landing_probe_n).

[0076] In some embodiments, the step of calculating the probability of outputting one or more affinity Normalizing the product of the probabilities of a reagent (probe) landing on a protein by a length coefficient. It may include. The length coefficient is such that longer proteins (for example, longer ones) are less long (for example, Compared to a shorter protein, a larger number of proteins can bind to it (for example, land on it). It is reasonable to take into account the assumption that affinity reagents are more likely to be randomly present. The number is a combination of n sets of cardinal numbers Len_i (representing the length of protein_i), that is, This can be expressed as a binomial coefficient "Len_i choose n", which can also be expressed as Choose(Len_i, n). It is possible. The length coefficient is a subset of elements of size n (for example, the number of elements that land on a protein). (A few probes) without considering their order, a set of elements Len_i (for example, of length i) It represents the number of different forms for selecting from proteins. Therefore, the length coefficient is normal. One or more affinity reagents (probes), either converted or split in this way, interact with the protein. The product of the probabilities of landing can be expressed as follows: [Product(P_landing_probe_1, P_landing_probe_2, …, P_landing_probe_n) / Choose(L en_i, n)]. This value is also the unnormalized probability of protein_i being present in the sample. It can also be called that.

[0077] The length of the candidate protein is related to the binding of the candidate protein to a specific affinity reagent. Recognize that this is an approximate value that approximates the number of available epitopes ("binding sites") in the location. In addition, in some embodiments, the process of calculating the probability of output is performed for each of the aforementioned probabilities. This is normalized to the total number of available binding sites in each of the candidate proteins. This may include, in some embodiments, the benefits of each of the candidate proteins The number of usable binding sites is determined experimentally using a qualitative process. In one embodiment, the qualitative process involves repeatedly binding an affinity reagent to a specific protein. The measurement is performed by doing so. In some embodiments, the qualitative process is described herein. The conditions shown during the aforementioned protein identification method and system are similar to or identical to those shown. It will be carried out under the following conditions.

[0078] In some aspects, the process of calculating the probability of output is not normalized. This may include normalizing the probability that protein i is present in the sample. Normalization is performed in the database. All normals across all proteins in the sample (e.g., multiple candidate proteins) This may include dividing by the sum of the probabilities that have not been processed. For example, all in the database All unnormalized values ​​across protein j (for example, multiple candidate proteins) The sum of probabilities is calculated as SUM(P(protein_j | probe[1, …, n], length(protein_j)) It can be expressed as follows. Therefore, the normalized probability that protein i is present in the sample is It can be expressed as follows: P(protein_i | probe[1, 2, …, n], length(protein_i)) = [Product(P_landing_p Lobe_1, P_Landing_Probe_2, …, P_Landing_Probe_n) / Choose(Len_i, n)] / SUM(P( Protein_j | Probe[1, …, n], Length(protein_j)))

[0079] In some aspects, the process of generating multiple probabilities involves multiple additional affinity reagents. This further includes repeatedly receiving additional information on the coupled measurements for each of the links. Each of the additional affinity reagent probes is one or more of several candidate proteins. It can be configured to selectively bind to candidate proteins. For example, the probability of outputting The first value is based on the two landing probes, as shown below, for each candidate tampon. Regarding the substance, it can be generated: P(protein_i | probe[1, 2], length(protein_i)) = [Product(P_landing_probe_ 1, P_landing_probe_2 / Choose(Len_i, 2)] / SUM(P(protein_j | probe[1, 2], Length (protein_j)))

[0080] Next, additional information on binding measurements for each of the multiple additional affinity reagent probes. The probability of being repeatedly received and output is repeatedly calculated as the next repeated value. It can be determined, thereby generating a second value for the probability of output. For example, the probability of output The second value of the rate is as shown below, based on the first two landing probes (probes 1 and 2) Based on the first two landing probes (probes 3 and 4), each candidate tamper Regarding the substance, it can be generated: P(protein_i | probe[1, 2, 3, 4], length(protein_i)) = [Product(P_landing_p Robe_1, P_Landing_Probe_2, P_Landing_Probe_3, P_Landing_Probe_4) / Choose(Len_ i, 4)] / SUM(P(protein_j | probe[1, 2, 3, 4], length(protein_j)))

[0081] In some embodiments, the output probability calculated or generated in step 115 is The binding measurement on the candidate protein generates the observed measurement outcome with a certain probability. The term "binding measurement outcome" is used herein to mean binding measurement. This refers to the information observed during the process. For example, the binding measurement outcome of an affinity reagent binding experiment. The substance may be either bound to or unbound to the reagent. In addition, or, multiple For each of the one or more candidate proteins in the candidate protein, The probability that a coupled measurement in quality does not produce the observed measurement outcome is calculated by computer. It can be calculated or generated by. In addition, or, in candidate proteins The probability that a combined measurement will produce an unobserved measurement outcome is calculated by computer. or may be generated. In addition, or, a series of binding measurements on candidate proteins. However, the probability of generating an outcome set can be calculated or generated by a computer. ru.

[0082] A "binding outcome set" is defined as a protein as used herein. This refers to multiple independent coupled measurement outcomes related to a series of parental outcomes from an experiment. Affinity reagent binding measurements can be performed on unknown proteins. This binding measurement includes the binding measurement outcome, and the set of all binding measurement outcomes. The bound outcome set is the combined outcome set. In some cases, the combined outcome set is This may be a subset of all observed combined outcomes. In some cases, Therefore, the set of coupled outcomes includes coupled metric outcomes that were not observed experimentally. That's fine.

[0083] In addition, or, one or more candidate proteins among multiple candidate proteins For each of these, the probability that the unknown protein is a candidate protein is calculated by computer It can be calculated or generated by [this method].

[0084] The probability in step 115 is the probability of all candidate outcomes for the unknown protein binding measurement. Based on comparisons with databases containing multiple protein sequences related to the protein, This can be achieved. Therefore, the input to the algorithm is a database of candidate protein sequences. , and binding measurements (for example, probes thought to have bound to an unknown protein) It may include a set. In some cases, the input to the algorithm is an arbitrary affinity reagent. The probability that something generates any binding measurement for any candidate protein (for example) Parameters related to estimating the trimer-level binding probability for each affinity reagent. - may include. The algorithm's output is the identity of the assumed candidate protein. Given the probability that a coupled measurement outcome or a set of coupled outcomes is observed, This may include. In addition, or the algorithm's output may include a coupled metric outcome. Alternatively, given a set of binding outcomes, for an unknown protein, candidate proteins The most likely identity to be selected from the set of characteristics, and that the identification is accurate. It may include a certain probability. In addition, or alternatively, the algorithm's output may be a high-probability candidate. The group of protein identities, and the unknown protein within the group It may include the probability that it is one of the proteins. If it is a protein, the probability of observing the binding measurement outcome can be expressed as follows: Possible: P (Binding measurement outcome | Protein)

[0085] In some embodiments, P (binding metric outcome | protein) is completely in silico It is calculated as follows. In some embodiments, P (binding measurement outcome | protein) is tan It is calculated based on or derived from the characteristics of the amino acid sequence of the protein. In this embodiment, P (binding measurement outcome | protein) is the amino acid sequence of the protein It is calculated independently of the findings. For example, P (binding measurement outcome | protein) is calculated independently of the data. Binding measurements were obtained in repeated experiments on isolates of candidate proteins, and P(binding) Measurement outcome | Protein) frequency: (total number of binding measurements divided by the number of outcomes) It can be determined experimentally by calculating from the number of binding measurements. In this context, P (binding measurement outcome | protein) is the past binding measurement for the protein. It is calculated based on or derived from the database. In some aspects P (binding measurement outcome | protein) is an unknown factor with censored binding measurement results. From a population of proteins, a reliable set of protein identifications is generated, and then, Among the set of unknown proteins that have been reliably identified as candidate proteins, It is calculated based on, or derived from, the frequency of the combined measured outcome.

[0086] In some embodiments, a population of unknown proteins is measured by P(binding outcome) | Tan It can be identified using the seed value of the protein, and the seed value is reliable for the candidate protein. Based on the frequency of the binding measurement outcome among the unidentified proteins that were matched, the results were improved. Obtain. In some embodiments, this process obtains updated combined measurement outputs. It is repeated using novel identifications generated based on the probability of the result, and then the novel The probability of the coupled measurement outcome is derived from an updated set of reliable identifications. This can be achieved. In some embodiments, binding measurement for one teria protein The parameters of the in silico model for predicting the probability of Utcam are reliable and the same Based on observed binding measurement outcomes in the unknown proteins being identified, learning and It will be updated. In some embodiments, this process will be updated This process is repeated using novel identifications generated based on the in silico model, and then, The probability of a new measured outcome may be generated from the updated in silico model. ru.

[0087] If the candidate protein is the protein being measured, then the binding measurement outcome is The probability of not observing can be expressed as follows: P(unbound outcome | protein) = 1 - P(bound outcome | protein)

[0088] If the candidate protein is the protein being measured, then N individual binding measurements The probability of observing a set of coupled measurement outcomes consisting of outcomes is equal to the probability of observing individual coupled measurement outcomes It can be expressed as the product of the probabilities for each of the Tocam: P(binding outcome set | protein) = P(binding measurement outcome 1 | protein) * P( Binding measurement outcome 2 | Protein) * … * P(Binding measurement outcome N | Protein)

[0089] Unknown protein is a candidate protein (protein) i The probability that ) is the same as the possible candidates This can be calculated based on the probability of the binding outcome set for each protein.

[0090] In some embodiments, an unknown protein is a candidate protein (protein i ) The probability is calculated by considering the binding outflow for each candidate protein j in the complete set of N candidate proteins. The muset is calculated as a fraction of the total probability of observation: TIFF2026076191000002.tif13144

[0091] In some embodiments, the binding measurement outcome set measures the binding of affinity reagent probes. Includes. In some embodiments, the binding measurement outcome set is non-affinity reagent probe. Includes specific binding.

[0092] In some embodiments, the method involves measuring all unknown proteins in the sample. The method further includes the step of applying the method to the subject. In some embodiments, the method may be one or more. For each of the candidate proteins, the candidate protein is one of the unknown proteins in the sample. The process further includes generating a confidence level that is consistent with the given values. The confidence level is a probability value. The confidence level may include the probability value of an error. The confidence level is a certain level of confidence (approximately 90%, 95%, 96%, 97%, 98%, 99%). , approximately 99.9% of, approximately 99.99% of, approximately 99.999% of, approximately 99.99999% of, approximately 99.999999% of, approximately 99.999999% , approximately 99.9999999, approximately 99.99999999, approximately 99.999999999, approximately 99.9999999999, approximately 99 A confidence level of 0.99999999999, approximately 99.9999999999999, approximately 99.99999999999999, or 99.999 It may include a range of probability values ​​that have an arbitrary confidence level of over 9999999999%.

[0093] In some embodiments, the method identifies proteins and their associated probabilities in a sample. For each unknown protein, the process involves independent production, as well as identification of the protein in the sample. The process further includes generating a list of all the unique proteins. In this method, in order to determine the amount of each candidate protein in the sample, a unique candidate ta The process further includes counting the number of identifications generated for each protein. In this context, the set of protein identification and associated probabilities shows high scores, high confidence, It can be filtered to include only identifications with low false detection rates and / or low false detection rates.

[0094] In some embodiments, the binding probability is related to the affinity reagent for the full-length candidate protein. It can be generated by... In some embodiments, the binding probability is that the protein fragment (e.g., complete...) Affinity reagents can be generated for (parts of) the entire protein sequence. For example, It's as if only the first 100 amino acids of each unknown protein are conjugated together. In this manner, if an unknown protein is processed and conjugated onto a substrate, the first 10 For epitope bindings other than 0 amino acids, all binding probabilities become zero, or The binding probability is set to a very low probability that represents the error rate, and the binding probability is set to that of the protein candidate. Each of these can be generated. A similar approach is to analyze the first 10 amino acids of each protein, 20 Amino acids, 50 amino acids, 100 amino acids, 150 amino acids, 200 amino acids, 300 amino acids, 400 amino acids It can be used when amino acids, or more than 400 amino acids, are conjugated to the substrate. A similar approach is used for the last 10 amino acids, 20 amino acids, 50 amino acids, 100 amino acids, 150 amino acids. amino acids, 200 amino acids, 300 amino acids, 400 amino acids, or more than 400 amino acids are incorporated into the base material. It can be used when jugated.

[0095] In some embodiments, the protein fragments before or after conjugation. If it may be processed to generate, the fragmentation of each protein is not definitive. It is not necessary. For example, proteins are physically sheared before conjugation to a substrate. This is possible. In such cases, the binding probability of the affinity reagent is the eye of the protein fragment. Dentities (e.g., the start and end points of a partial sequence of a complete protein candidate containing the fragment) It can be modeled using stop points as joint models. For example, the expectation maximization approach can be used for each This approach can be used to generate binding probabilities for protein candidates, and the observed Based on the binding measurements, the most likely fragment generated by the protein candidate was identified. The hypothesis is iteratively refined, and then each affinity trial for the modeled protein fragment is performed. Update the probability of drug binding.

[0096] In some cases, modeling protein fragments from a specific protein candidate. Prior knowledge about the likelihood of generating a fragment may be incorporated. For example, protein fragments Prior knowledge about the expected length distribution may be given. As another example, lysine or This provides favorable prior knowledge for protein fragments sandwiched by arginine, while untreated proteins The quality is given when the substance is treated with trypsin enzyme before conjugation. In some embodiments, the binding measurement is compared to that of a candidate protein sequence. The database may contain protein fragments. For example, trypsin from a source sample. When a peptide mixture derived from a compound is conjugated onto a substrate, the protein candidate lithography The complete protein sequence is generated from the in silico digestion of an unprocessed protein sequence database. It may contain all peptides that have been completely trypsin-digested. In such cases, The results from affinity reagent binding measurements for each unknown protein fragment in the sample This can be used to identify the most likely trypsin-digested peptide. In cases like these, the resulting peptide identity and / Alternatively, the quantity can be converted to a protein-level measurement using a protein estimation approach. Indeed, numerous examples of this approach exist, for example, in the field of mass spectrometry.

[0097] In some embodiments, one candidate protein is used as one half of a matching pair, while an unknown protein is used. If it cannot be assigned to the protein, it may be a potential protein candidate as one half of a matching pair. The supplementary group can be assigned to the unknown candidate. The confidence level is the protein in the group. It can be assigned to one of the candidate unknown proteins. The confidence level is, It may include probability values. Or, the confidence level may include probability values ​​that are incorrect. Or The confidence level is a certain level of confidence (for example, approximately 90%, approximately 95%, approximately 96%, approximately 97%, approximately 98%). It may include a range of probability values ​​that have, arbitrarily, a confidence level of approximately 99%. For example, The unknown protein could be a strong fit for two candidate proteins. These can have high sequence similarity (e.g., protein isoforms, canonical sequences). (A protein with a single amino acid mutation compared to the others). In these cases, the individual proteins While it is possible that a candidate for the drug may not be assigned with high confidence, The confidence level is among the "protein groups" that include two strongly matched protein candidates. This could be attributed to an unknown protein, consistent with a single, but unknown, member.

[0098] In some embodiments, to detect a state in which an unknown protein is not optically distinguishable, This may require considerable effort. For example, in rare cases, two or more proteins may be used as a base material. The possibility of joining the same "well" or position, and the effort required to prevent this from happening. Nevertheless, there are cases where conjugated proteins are non-specific. The samples may be treated with different dyes, and the signals from the dyes may be measured. In situations where the above proteins are not optically distinguishable, the signals generated from the pigment The bond can be stronger than the bond between a single protein and multiple proteins. It can be used to indicate the location of something.

[0099] In some embodiments, multiple candidate proteins are found to be samples of an unknown protein. Human or organismal DNA or RNA obtained from or derived therefrom, and sequencing It is generated or modified through analysis.

[0100] In some aspects, the method derives information about post-translational modifications of an unknown protein. This further includes the process of extracting information. Information about post-translational modifications is based on knowledge about the nature of specific modifications. It may include the existence of post-translational modifications. The database can be considered as the exponential product of PTMs. For example, if a candidate protein sequence is assigned to an unknown protein, it will be assayed. The affinity reagent binding pattern for the protein was found to be different from previous experiments for the same candidate. Databases including binding measurements of affinity reagents can be compared. For example, data on binding measurements Tabes is a nucleic acid programmer that contains unmodified proteins with known sequences at known locations. Binding to a Nucleic Acid Programmable Protein Array (NAPPA) It can be derived from that.

[0101] Alternatively, a binding measurement database can be used to determine if a candidate protein sequence is unknown for a protein. It may be derived from a previous experiment assigned with reliability. Assayed protein and Discrepancies in combined measurements with existing measurement databases affect the likelihood of post-translational modifications. It can provide information on the subject. For example, if affinity agents are listed in the database as candidate substances, It has a high frequency of binding to proteins, but does not bind to the assayed protein, There is a higher likelihood that post-translational modifications exist somewhere on the protein. For affinity reagents with binding mismatch, if the binding epitope is known, post-translational modification is performed. The location may be at or near the location of the affinity reagent's binding epitope. Yes. In some aspects, information about specific post-translational modifications is available for specific post-translational modifications. Before and after treating the protein-substrate conjugate with an enzyme that specifically removes the compound, This can be derived by performing repeated affinity reagent measurements. For example, binding measurement The standard is that, with respect to a series of affinity reagents, they are obtained before processing the substrate with phosphatase. This can then be repeated after treatment with phosphatase. It binds to an unknown protein before treatment, but does not bind after phosphatase treatment (difference Affinity reagents for certain bonds provide evidence of phosphorylation. Affinity reagents for different bonds Therefore, if the recognized epitope is known, phosphorylation is performed on the bound epitope to the affinity reagent. It may be located at or near the location of a pitope.

[0102] In some cases, the number of a particular post-translational modification is related to the affinity for that particular post-translational modification. This can be determined using reagent-based binding measurements. For example, recognizing phosphorylation events. Antibodies may be used as affinity reagents. The binding of this reagent to an unknown protein is... This may indicate the presence of at least one phosphorylation. In some cases, an unknown protein The number of distinct post-translational modifications of a particular type in a substance is specific to a particular post-translational modification. This can be determined by counting the number of binding events measured for a particular affinity reagent. For example, a phosphorylation-specific antibody may be conjugated to a fluorescent reporter. In this case, the intensity of the fluorescence signal indicates the phosphorylation-specific affinity reagent bound to the unknown protein. It can be used to determine the number of phosphorylation-specific affinity bound to an unknown protein. The number of reagents is then used to determine the number of phosphorylation sites in the unknown protein. It may be used. In some embodiments, a more accurate number, identification, or arrangement of post-translational modifications may be used. To derive this, evidence from affinity reagent binding experiments suggests the possibility of post-translational modification. Existing knowledge about certain amino acid sequence motifs or specific protein locations ( For example, it can be combined with dbPTM, PhosphoSitePlus, or UniProt. For example, if the position of post-translational modifications cannot be precisely determined solely from affinity measurements, then the position of interest is... Positions containing amino acid sequence motifs often associated with post-translational modifications in elephants may be advantageous. ru.

[0103] In some embodiments, the step of generating probabilities is related to the information of the coupled measurement, and the detector This includes taking into account the error rate of the detector. The error rate of the detector may include the true landing rate. If so, the detector error rate can be due to the failure of the probe to "land" on the protein, and this is For example, if a probe gets stuck in the system and is not properly flushed out, Or the probe was unexpected based on previous qualitative and testing of the probe. This is the case when it binds to a protein. Alternatively, the error rate of the detector is the physical error of the detector. This may be caused by one or more detectors used to obtain information on coupled measurements. The following can be obtained from the specifications. The detector error rate may include one or more of the following: The physical error rate of the detector, the off-target coupling rate, or errors due to stacked probes. Error rate. In some embodiments, the error rate of the detector is set to the estimated error rate of the detector. Alternatively, the detector's prediction error rate can be set by the computer user. In some aspects, the detector's estimated error rate is approximately 0.0001, approximately 0.0002, approximately 0.0003, Approximately 0.0004, approximately 0.0005, approximately 0.0006, approximately 0.0007, approximately 0.0008, approximately 0.0009, approximately 0.001, approximately 0.002, approximately 0.003, approximately 0.004, approximately 0.005, approximately 0.006, approximately 0.007, approximately 0.008, approximately 0.009, approximately 0.01, approximately 0.02, approximately 0 It is 0.03, approximately 0.04, approximately 0.05, approximately 0.06, approximately 0.07, approximately 0.08, approximately 0.09, approximately 0.1, or greater than approximately 0.1. .

[0104] A hit table is a table where each column of the hit table contains a different protein (for example, different proteins). (having a length) and / or each row of the hit table is different It can be generated to represent the probe. A given element of the hit table (for example, row j) In column i), each value represents the reaction of a given probe j to a given protein after exposure to the sample. It may include a value indicating whether or not it can bind to i. For example, if probe j binds to protein i. If possible, elements of the hit table can be set to 1 (for example, row j and column j). In i), and otherwise it can be set to 0. This information is gradually Since it is possible to reach it, the hit table may be calculated iteratively.

[0105] A probability matrix can be calculated or generated from the hit table. Probe j is the sample If exposed to protein i inside, then each value of a given element in the probability matrix is: This may include a value indicating the probability that the measurement will be observed. This probability is P(protein_i | probe_j) It can be expressed as follows: The corresponding hit table entry is greater than 1. If either or equal to 1, the probability matrix entry is the true landing rate (e.g., P_Landing_ It can be set to probe_j). If the corresponding hit table entry is 0 In some cases, the probability matrix entry is set to the detector's error rate (e.g., 0.0001). It can be done. The error rate of the detector may include one or more of the following: detection The physical error rate of the instrument, the off-target coupling rate, or the error rate due to stacked probes. .

[0106] In some embodiments, the process of repeatedly generating multiple probabilities is carried out from subsequent iterations, Further comprising removing one or more candidate proteins from the candidate protein, This reduces the number of iterations required to perform the iterative generation of probabilities. In some embodiments, removing one or more candidate proteins is a candidate protein It is based at least on predetermined standards for binding measurements related to protein. In some embodiments The predetermined criteria are that the first of multiple affinity reagent probes falls below a predetermined threshold. It contains one or more candidate proteins having rotational binding measurements. The protein is, for example For example, after measuring the binding of k probes, the P(protein i | probe [1..k]) If the value is less than 0.01, less than 0.001, less than 0.0001, less than 0.00001, less than 0.000001, or less than 0.0000001 If the condition is met, it may be excluded from consideration. Proteins can also be experimentally removed from the sample. If they are removed from consideration, they may be excluded from consideration.

[0107] In some embodiments, as described elsewhere in this specification, the probability of This is normalized with respect to the length of the candidate protein. In some embodiments, this specification As described elsewhere in the book, each of the probabilities is the probability of multiple candidate proteins. It is normalized with respect to the sum of 10 or less, 20 or less, 30 or less, 40 or less, 50 or less, 60 or less, 70 or less, 80 or less, 90 or less, 100 or less, 150 or less, 200 or less, 250 or less, 300 or less, 350 or less, 400 Includes 100 or fewer, 450 or fewer, 500 or fewer, or more than 500 affinity reagent probes.

[0108] In some embodiments, probabilities are repeatedly generated until a predetermined condition is met. In some embodiments, the given conditions are at least 50%, at least 55%, and at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, At least 90%, at least 91%, at least 92%, at least 93%, at least 94%, less At least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or a small This includes generating each of multiple probabilities with at least 99.9% confidence.

[0109] In some embodiments, the method identifies one or more unknown proteins in a sample. The process further includes generating paper or electronic reports. The test determines that, for each candidate protein, the candidate protein is present in the sample. The confidence level can be further indicated. The confidence level may include probability values. Alternatively, the confidence level may be, It may include probability values ​​that are incorrect. Alternatively, the confidence level may be a certain level of confidence (e.g., 90%). A probability within a range of values, with any of the following confidence levels: 95%, 96%, 97%, 98%, or 99%. The value may include. Paper or electronic reports may have a threshold for the expected false detection rate (for example). Less than 10%, less than 5%, less than 4%, less than 3%, less than 2%, less than 1%, less than 0.5%, less than 0.4%, less than 0.3% A list of protein candidates identified with a false detection rate of less than 0.2% or less than 0.1%. This can be further demonstrated by sorting protein identifications first in descending order of confidence, with the false detection rate being the first factor. This can be inferred by the fact that, at any point in the sorted list, The false detection rate can then be calculated as 1 - avg_c_prob, where avg_c_prob is a list. For all proteins, at the current point in time or earlier (with higher confidence) This is the average candidate probability. The list of protein identifications that fall below the desired false detection threshold is: Subsequently, in the sorted list, from the earliest point where the false detection rate is higher than the threshold... It can be generated by returning all previous protein identifications. Alternatively, the desired The list of protein identifications that fall below the false detection threshold is the list of false detections in the sorted list. The detection rate is below or equal to the desired threshold, including the latest points. It can be produced by returning all the previous proteins.

[0110] In some aspects, the sample includes a biological sample. The biological sample is obtained from the subject. It may be done. In some embodiments, the method is based on at least several probabilities, and the target The process further includes identifying a disease state or disorder in the present. In some embodiments, The method involves counting the number of identifications generated for each protein candidate. This further includes a process for quantifying quality. For example, the absolute amount of protein present in the sample (tan). The number of protein molecules counts the number of reliable identifications generated from protein candidates. It can be calculated by the following. In some embodiments, the amount is assayed It can be calculated as a percentage of the total number of unknown proteins. In some aspects Therefore, the raw identification number is used to eliminate system errors from the instrument and detection system. It can be calibrated. In some embodiments, the amount can be adjusted depending on the variability in the detectability of the protein candidate. To eliminate the bias in the amount caused by this, it can be calibrated. The detectability of the protein is It can be evaluated through experimental measurements or computer simulations.

[0111] Diseases or disorders include infectious diseases, immune disorders or autoimmune diseases, cancer, genetic disorders, and degenerative diseases. This could be a disease, lifestyle-related disease, wound, rare disease, or age-related disease. Infectious diseases are It can be caused by bacteria, viruses, fungi, and / or parasites. Limited examples include bladder cancer, lung cancer, brain cancer, melanoma, breast cancer, non-Hodgkin lymphoma, and children. Cervical cancer, ovarian cancer, colon cancer, rectal cancer, pancreatic cancer, esophageal cancer, prostate cancer, kidney cancer, This includes skin cancer, leukemia, thyroid cancer, liver cancer, and uterine cancer. Hereditary disorders or genetic conditions. Some examples of transmissible disorders, though not limited to them, include cystic fibrosis and Charcot-Magnesia. Lee-Tooth disease, Huntington's disease, Peutz-Jeghers syndrome, Down syndrome, arthritis This includes rheumatism and Tay-Sachs disease. Non-limited examples of lifestyle-related diseases include obesity, diabetes, Arteriosclerosis, heart disease, stroke, hypertension, cirrhosis, nephritis, cancer, chronic obstructive pulmonary disease (COPD) This includes hearing problems and chronic back pain. Examples of wounds are not limited to these. However, abrasions, brain injuries, contusions, burns, concussions, congestive heart failure, injuries at construction sites, dislocations Chest instability, fracture, hemothorax, herniated disc, hip pointer, hypothermia, laceration, nerve compression Distress, pneumothorax, rib fracture, sciatica, spinal cord injury, wounds to tendons, ligaments, and fascia, trauma This includes brain injury and whiplash.

[0112] In another context, in this specification, candidate proteins in a sample of an unknown protein are A computer-based method for identification is disclosed. The method involves the sample... Binding measurements for each of multiple affinity reagent probes to an unknown protein. The process may include receiving information via a computer. The affinity reagent probe is k-ma - may be an affinity reagent probe. In some embodiments, k-mer affinity reagent probe Each of the steps is selective for one or more candidate proteins from among multiple candidate proteins. It is set to bind to [a certain protein]. The information from the binding measurement suggests that it has bound to an unknown protein. It may include a set of probes.

[0113] Next, at least a portion of the binding measurement information is in a database containing multiple protein sequences. Each protein sequence can be compared by a computer. It could correspond to one candidate protein among the candidate proteins. Multiple candidate proteins are At least 10, at least 20, at least 30, at least 40, at least 50, At least 60, at least 70, at least 80, at least 90, at least 100 , at least 150, at least 200, at least 250, at least 300, at least 350, at least 400, at least 450, at least 500, at least 600, fewer At least 700, at least 800, at least 900, at least 1000, or more than 1000 This may include different candidate proteins.

[0114] Next, one or more candidate proteins from multiple candidate proteins are further investigated. It can be removed (for example, from subsequent calculations, iterations, calculations, or probability generation). Removal of one or more candidate proteins from multiple candidate proteins is important for binding measurement. The report can be based at least on comparisons with databases containing multiple protein sequences. .

[0115] In some embodiments, the removal of one or more candidate proteins is performed by the candidate proteins It is based at least on predetermined criteria for binding measurements related to the above. In some embodiments, The predetermined criterion is the adhesion to the first of several affinity reagent probes below a predetermined threshold. It comprises one or more candidate proteins having a combined measurement. In some embodiments, Candidate proteins are, for example, identified after the binding of k probes is measured, and their P(protein) Quality i | Probe [1..k]) is less than 0.01, less than 0.001, less than 0.0001, less than 0.00001, 0.000001 If the value is less than 0.0000001, it may be excluded from consideration. Proteins can also be excluded from consideration. Even if it is experimentally removed from the sample, it may be excluded from consideration.

[0116] In some embodiments, the number of affinity reagent probes is 10 or less, 20 or less, or 30 or more. Lower, 40 or less, 50 or less, 60 or less, 70 or less, 80 or less, 90 or less, 100 or less, 150 Less than 200 pieces, 250 pieces or less, 300 pieces or less, 350 pieces or less, 400 pieces or less, 450 pieces or less, 500 pieces or less Includes, or more than 500, affinity reagent probes.

[0117] In some embodiments, affinity reagent probes on which binding measurements are performed are It is determined completely before the measurement is carried out. In some embodiments, the combined measurement related thereto The set or sequence of affinity reagent probes to be determined is based on the results obtained up to that point. Based on repeated computer analysis of the combined measurements, the results may be modified during the experiment or This can be derived. For example, the ordering of affinity probes is for unidentified unknown proteins. Therefore, we will prioritize binding experiments using probes that are more likely to produce clear identification. Therefore, it can be iteratively optimized. Such optimization involves identifying unknown proteins that were not identified until then. Regarding quality, the top two, top three, top four, top five, or the best This could be based on selecting probes that distinguish between the top five or more candidate protein sequences. .

[0118] In some embodiments, the method identifies one or more unknown proteins in a sample. The process further includes generating paper or electronic reports. The test determines that, for each candidate protein, the candidate protein is present in the sample. The confidence level can be further indicated. The confidence level may include probability values. Alternatively, the confidence level may be, It may contain a probability value that is incorrect. Alternatively, the confidence level may be a certain level of confidence (e.g., 90%). A range of probability values ​​having any of the following confidence levels: 95%, 96%, 97%, 98%, and 99%. , may include. In some aspects, the sample includes a biological sample. A biological sample is, It may be obtained from the subject. In some embodiments, the method is based on at least multiple probabilities. The process further includes identifying the disease state or impairment in the subject.

[0119] Diseases or disorders include infectious diseases, immune disorders or autoimmune diseases, cancer, genetic disorders, and degenerative diseases. This could be a disease, lifestyle-related disease, wound, rare disease, or age-related disease. Infectious diseases are It can be caused by bacteria, viruses, fungi, and / or parasites. Limited examples include bladder cancer, lung cancer, brain cancer, melanoma, breast cancer, non-Hodgkin lymphoma, and children. Cervical cancer, ovarian cancer, colon cancer, rectal cancer, pancreatic cancer, esophageal cancer, prostate cancer, kidney cancer, This includes skin cancer, leukemia, thyroid cancer, liver cancer, and uterine cancer. Hereditary disorders or genetic conditions. Some examples of transmissible disorders, though not limited to them, include cystic fibrosis and Charcot-Magnesia. Lee-Tooth disease, Huntington's disease, Peutz-Jeghers syndrome, Down syndrome, arthritis This includes rheumatism and Tay-Sachs disease. Non-limited examples of lifestyle-related diseases include obesity, diabetes, Arteriosclerosis, heart disease, stroke, hypertension, cirrhosis, nephritis, cancer, chronic obstructive pulmonary disease (COPD) This includes hearing problems and chronic back pain. Examples of wounds are not limited to these. However, abrasions, brain injuries, contusions, burns, concussions, congestive heart failure, injuries at construction sites, dislocations Chest instability, fracture, hemothorax, herniated disc, hip pointer, hypothermia, laceration, nerve compression Distress, pneumothorax, rib fracture, sciatica, spinal cord injury, tendon, ligament, fascia injury, trauma This includes brain injury and whiplash.

[0120] In some aspects, the method involves small molecules (e.g., metabolites) rather than proteins. This includes the steps of identifying and quantifying glycans. For example, sugars with diverse properties. Alternatively, affinity reagents such as lectins or antibodies that bind to a combination of sugars can be used to bind to glycans. May be used to identify affinity for various sugars or sugar combinations. The properties of the reagent can be characterized by analyzing its binding to commercially available glycan arrays. The unknown glycan is used in a hydroxyl group-reactive chemical reaction to create a functionalized substrate. It can be conjugated, and binding measurements can be obtained using affinity reagents that bind to glycans. It is possible. Measuring the binding affinity of an affinity reagent to an unknown glycan on a substrate, which has a specific sugar, Alternatively, it can be used to directly quantify the number of glycans that have a specific combination of sugars. To obtain, or to identify the structure of each unknown glycan, one or more bonds The combined measurement is performed using the estimation algorithm described herein to determine the candidate glycan structure. The combined measurement can be compared with the one predicted from the database. In some embodiments, The protein is bound to the substrate, and the binding measurement using a glycan affinity reagent is performed on the protein. It is generated to identify the glycans bound to the protein. Furthermore, the binding measurement is performed on the protein. To generate identification of kubon sequences and conjugated glycans in a single experiment. This can be done using both glycan affinity reagents and protein affinity reagents. Another example is... Common metabolites include sulfhydryls, carbonyls, amines, or active hydrogens. A functionalized substrate is created using a chemical reaction that targets the coupling group found therein, and metabolites These can be conjugated. Binding measurements are performed on specific functional groups, structural motifs, or metabolites. Alternatively, this may be done using affinity reagents with different properties. The result obtained is The binding measurement can be compared with the predicted binding measurement for the candidate small molecule database, and The estimation approach described herein identifies metabolites at each location on the substrate. It can be used for that purpose.

[0121] Computer-controlled system This disclosure describes a computer system programmed to perform the methods described herein. Provided. Figure 2 shows computer system 201, which: unknown protein in sample To receive information on the binding measurement of affinity reagent probes to the quality, the binding measurement information is used as a candidate. To compare against a database containing multiple protein sequences corresponding to the coprotein To repeatedly generate the probability that the candidate protein is present in the sample. It is programmed to do so, or is configured to do so under other circumstances.

[0122] Computer system 201 can, for example, determine affinity levels for unknown proteins in a sample. The process involves receiving information on drug probe binding measurements, and then assigning the binding measurement information to the candidate protein. The process involves comparing against a database containing multiple protein sequences, and / or candidate sequences. The methods of this disclosure, such as the process of repeatedly generating the probability of the coprotein being present in the sample, Furthermore, it allows control over various aspects of the system.

[0123] Computer system 201 remotely accesses the user's electronic device or electronic device. It could be a computer system located at a specific location. An electronic device could be a portable electronic device. Computer system 201 is a central processing unit (CPU, also referred to as "processor" in this specification). And "computer processors") 205, which are single-core or multi-core. This can be a core processor or multiple processors for parallel processing. System 201 also has memory or memory location 210 (for example, random access (flash memory, read-only memory, flash memory), electronic storage unit 21 5 (e.g., hard disk), communication input for communication with one or more other systems Surface 220 (e.g., network adapter), as well as cache, and other notes. Lee, data storage and / or peripheral devices such as electronic display adapters 22 Includes 5. Memory 210, storage unit 215, interface 220, and peripheral devices. The 225 communicates with the CPU 205 via a communication bus (solid wire) on the motherboard, etc. The unit 215 is a data storage unit (or data storage) for storing data. Possible. Computer system 201 provides assistance to communication interface 220. In response, it may be functionally connected to the computer network ("Network") 230. Network 230 is the Internet, an intern et) and / or extranets, or the Internet, and communication This could be an intranet and / or extranet. Network 230 is In some cases, it is a telecommunications and / or data network. Work 230 enables distributed computing such as cloud computing. Network 230 may include one or more computer servers. In such cases, with the assistance of computer system 201, computer system 201 A peer that can enable a connected device to function as either a client or a server. A top-down network can be implemented.

[0124] CPU 205 can be embodied in the form of a program or software, machine-readable A sequence of instructions can be executed. The instructions are in memory such as memory 210. It may be stored in the application. The instruction may be directed to CPU 205, which then opens up To carry out the method shown, CPU 205 can be programmed or otherwise configured. Examples of tasks performed by PU 205 include fetching, decoding, executing, and writing back. It may include.

[0125] CPU 205 may be part of a circuit such as an integrated circuit. One or more other components of System 201 The components of the circuit may be included in the circuit. In some cases, the circuit is a collection of components for a specific purpose. It is an integrated circuit (ASIC).

[0126] Storage unit 215 stores drivers, libraries, and saved programs. It can store files such as the following. The storage unit 215 can store, for example, user files. It can store user data such as references and user programs. The user system 201, in some cases, is an intranet or internet A remote server that communicates with computer system 201 via the Internet One or more additional data located outside of computer system 201, such as those located outside of the computer system 201. It may include a storage unit.

[0127] Computer system 201 connects to one or more remote locations via network 230. It can communicate with computer systems. For example, computer system 201, It can communicate with the user's remote computer system. Examples of systems include personal computers (such as portable PCs), slates, or tablets. Redline PCs (for example, Apple® iPad, Samsung® Galaxy Tab), phones Devices, smartphones (for example, Apple® iPhone, Android devices, Bluetooth devices) Including ackberry®, or personal digital assistants. Users can Computer system 201 is accessible via network 230.

[0128] Methods described herein include, for example, memory 210 or electronic storage On the electronic storage locations of the computer system 201, such as on page unit 215 By stored, machine-executable code (e.g., computer processor), It can be performed. Machine-executable or machine-readable code is software. It can be provided in this form. During use, the code can be executed by processor 205. How many In that case, the code may be retrieved from the storage unit 215, and the process For quick access to the data 205, it can be stored in memory 210. In some situations The electronic storage unit 215 may be excluded, and machine-executable instructions are in memory 210. It will be stored in memory.

[0129] The code may be precompiled and run on a processor adapted for that purpose. It can be configured for use on machines that have it, or it can be compiled during runtime. The code is either precompiled or compiled at runtime (as-compiled). A programming language that can be chosen to allow the execution of code in the ) style. It may be provided.

[0130] Systems and methods provided herein, such as computer system 201. This aspect can be embodied in the form of programming. Various aspects of technology can be read by machines. Transported or materialized by one of the available media, typically by machine (or p "Items" are the form of executable code and / or associated data (in a processor). It may be considered as "a product" or "a machine executable code". (For example, read-only memory, random access memory, flash memory) or It can be stored on an electronic storage unit, such as a disk. The medium is a computer, processor, or the tangible memory of a similar device, or This includes various semiconductor memory, tape drives, disk drives, and similar devices. This may include any or all of the related modules, and these are software It can provide non-transient storage at any time during programming. All of the software Or part of it, the Internet or various other telecommunications networks Communication may occur occasionally via the network. Such communication may occur, for example, through a single computer. Transferring data from one processor to another, for example, a management server or host. From a computer to an application server computer platform, It may be possible to load software. Therefore, it may be possible to carry software elements. Another type of medium is used by passing through a physical interface between local devices. things, things used via wired and optical fixed telephone networks, and This includes light waves, radio waves, and electromagnetic waves, such as those used through various air links. . Wired or wireless links, optical links, or similar, which carry such waves. Physical elements may also be considered as a medium for carrying software. Unless limited to non-transient and tangible "storage" media, as used, computers In a processor or machine, words like "readable medium" refer to a medium that is not readable by the processor for execution. This refers to any medium involved in providing instructions.

[0131] Therefore, machine-readable media such as computer-executable code are This includes, but is not limited to, tangible storage media, carrier media, or physical transmission media. It can take many forms, including non-volatile storage media such as any computer. Any storage device in a printer or similar device, or the data shown in the drawings. This includes optical or magnetic disks, such as those that may be used to implement databases, etc. Volatile storage media include the main memory of computer platforms, etc. This includes dynamic memory. The tangible transmission medium is the bus in the computer system. Coaxial cables, including wiring; copper wires, and optical fibers. The carrier wave transmission medium is , in the form of electrical signals or electromagnetic signals, or in radio frequency (RF) and infrared (IR) signals. It may take the form of sound waves or light waves, such as those generated during data communication. Common forms of media readable by a printer include, for example, the following: Flexible disks, hard disks, magnetic tapes, and any other magnetic tapes. Media, CD-ROM, DVD or DVD-ROM, any other optical media, punched card paper tape, perforated Any other physical storage medium having the pattern, RAM, ROM, PROM and EPROM, F LASH-EPROM, any other memory chip or cartridge, carries data or instructions A carrier wave, a cable or link that carries such a carrier wave, or a computer that reads Any other medium from which programming code and / or data may be incorporated. Many of these forms of computer-readable media are processed by a processor. It may be involved in carrying one or more sequences of one or more instructions.

[0132] Computer system 201, for example, algorithms, combined measurement data, candidate tans User interface for providing user selection of proteins and databases An electronic display 235 including (UI)240 may be included, or communicates with it. This is possible. Examples of UI are not limited to graphical user interfaces. Includes (GUI) and web-based user interfaces.

[0133] The methods and systems disclosed herein may be performed by one or more algorithms. The algorithm is executed by software during execution by the central processing unit 205. The algorithm obtains, for example, affinity reagent probes to an unknown protein in the sample. It can receive information on binding measurements, and the information on binding measurements corresponds to the candidate protein. It can be compared against databases containing multiple protein sequences, and / or The probability of a candidate protein being present in a sample can be repeatedly generated. [Examples]

[0134] Example 1 - Protein identification using a database of 6 candidate proteins The database contains six candidate proteins of length {276, 275, 151, 437, 244, 644}. Let's examine the situation. In addition, the experiment showed that each of them had a 25% likelihood of binding to a given trimer. This will be performed using five probes. The other trimers to which these reagents bind are data It is not found in any protein in the base.

[0135] A hit table is constructed for each probe in the database. (Rows = Probe #1~#5, Columns = SEQ ID 1~6) TIFF2026076191000003.tif33128

[0136] In particular, this information arrives gradually, and therefore can be calculated iteratively. As shown, from the hit table, P(protein_i | probe_j) is, Evaluated to generate a hit table. For a given entry, hit table >= 1 If so, P_landing_probe_n = true landing rate = 0.25 is used; otherwise, Please note that if the table value is 0, then P (detector error) = 0.0001 will be used. . TIFF2026076191000004.tif39128

[0137] Note that many cells contain a probability of 0.0001. This small probability is due to the detector error. This could be the cause.

[0138] First, the unnormalized probability of each protein is given for each candidate protein. It is calculated as the product of: TIFF2026076191000005.tif3144

[0139] Next, the length normalization is calculated, which means that several probes are given a protein The number of patterns that land on the quality is expressed as a function of the protein length. Length normalization is done by choosing It is expressed as the word (Len_i, n). For example, the first protein is the length of [276 choose 5] The first protein has a normalization of length, and the second protein has a normalization of length [275 choose 5]. In some embodiments, the normalization of length is calculated as Len_i! / (len_i! - n!) It can be calculated as the number of columns, where ! represents the factorial. TIFF2026076191000006.tif3131

[0140] Next, the product (ProductP) from the above is normalized to take this length correction into account. It is normalized by dividing by the formula, and this is shown below: TIFF2026076191000007.tif3144

[0141] Next, the probability sets across the entire database so that they sum up to 1. This is normalized. This is done by summing them up so that the LenNormP value is 1.53E-13, and then the final To achieve a balanced probability, we divide each of LenNormP by this normalization. This is achieved by: TIFF2026076191000008.tif3139

[0142] While four of the proteins are extremely unlikely, the distinction between proteins 1 and 2 is clear. It should be noted that this is somewhat difficult. From the database, this is Tampa This is presumably because there is only one difference in point deletion between protein 1 and protein 2. Proteins 2 and 3 each share a 50% chance, while proteins 3-6 essentially have a zero chance. I want to attract attention for what I do.

[0143] In experimental techniques, probes are detected sequentially; therefore, this function is iterative. It is desirable to perform the calculation in this way. There are several different methods to achieve this. An example of this is shown below.

[0144] Example 2 - Protein identification using an antibody mixture In accordance with the disclosed aspects, the identification of 1,000 unknown human proteins was carried out by Santa Cruz Biology. We decided to obtain a binding assay using a pool of commercially available antibodies from the otechnology catalog. Therefore, it was conducted as a benchmark test. 1,000 unknown proteins were found to be approximately 21,005 Randomly selected from the Uniprot protein database, which contains individual human proteins. It is reactive to proteins and is available from the Santa Cruz Biotechnology catalog. A list of available monoclonal antibodies was downloaded from an online antibody registry. This list contained 22,301 antibodies, and the Uniprot Human Protein Database The list was filtered to include 14,566 antibodies matched to the proteins in the sample. The complete population of antibodies modeled in the experiment contained these 14,566 antibodies. The experimental evaluation of the binding of antibody mixtures to 1,000 unknown protein candidates is as follows: It was carried out in this manner.

[0145] First, 50 antibody mixtures were modeled. 5,000 antibodies were randomly selected from the total antibody population.

[0146] Next, for each mixture, the binding probability for any of the unknown proteins in the mixture is... It was decided that the goal of the protein was to estimate their identity. In that sense, it is "unknown," but the algorithm is the true nature of each "unknown protein." Please note that the identity is known. The mixture is against the unknown protein. If the mixture contains an antibody that binds to the unknown protein, it is assigned a binding probability of 0.99. If the corresponding antibody was not present, a binding probability of 0.0488 was assigned.

[0147] The probability of nonspecific binding for a mixture is that any individual antibody binds to proteins other than its target. Based on the expected probability of binding to the protein and the number of proteins in the mixture, the model It was modified. Regarding the evaluation by this experiment, individual antibodies were modified to something other than their target protein. The probability of a nonspecific binding event, namely binding, being 0.00001(1E-5), was considered to be 0.00001(1E-5). The probability of a nonspecific binding event for an antibody mixture is that any one antibody in the mixture This is the probability of nonspecific binding. This probability ranges from 1 to the probability that all 5000 antibodies in the mixture bind. Subtracting the probability of non-specific binding, that is, 1 - (1 - 1e-5)^1000 = 0.0488 and And so, it was calculated.

[0148] For each unknown protein, binding is related to the probability of the antibody mixture binding to the unknown protein. Based on this, each of the measured antibody mixtures was evaluated. Minimum value of 0 and maximum value of 1. A uniform distribution with values ​​is randomly sampled, and the resulting number is If the probability of the antibody mixture binding to the unknown protein is less than the probability of the mixture binding to the unknown protein, the experiment will be considered for the mixture. A binding event occurred. Otherwise, the experiment would have resulted in a non-binding event for the mixture. This resulted in the following protein prediction using all the binding events evaluated: It will be implemented on [date].

[0149] For each unknown protein, a series of evaluated binding events (50 in total, one in the mixture) (1 per month) for each of the 21,005 protein candidates in the Uniprot database It was evaluated. More specifically, the probability of observing a series of coupling events was calculated for each candidate. The probability was determined. The probability is calculated for each individual mixture across all 50 mixtures measured. It was calculated by multiplying the probability of a join event by the probability of a non-join event. The join probability is, The probability of non-joining is calculated in the same format as described above, and is obtained by subtracting the probability of joining from 1. Therefore, the protein query candidate with the highest binding probability is the unknown protein. This is the presumed identity. It is not certain that the identification is accurate for each individual protein. The rate is calculated as the probability of the top-ranking individual candidate, divided by the sum of the probabilities of all candidates. It was done.

[0150] Using the estimated identity for each of the 1,000 unknown proteins The unknown proteins were sorted in descending order of their identification probability. Tooff is defined as the percentage of inaccurate identifications among all preceding identifications in the list being 1%. As such, they were selected. Overall, out of 1,000 unknown proteins, 551 were selected, representing 1%. Identified with an inaccurate identification rate.

[0151] Example 3: Protein identification using binding measurement outcomes The methods described herein involve the binding of affinity reagents to unidentified proteins and / or can be applied to different subsets of data related to unjoined data. Several aspects In this specification, the method described herein is used to identify among the measured combined outcomes. This can be applied to experiments where a subset of the outcome is not considered (e.g., unbound metric outcomes). These methods, which do not consider a subset of the defined combined outcomes, are described in this specification. In the book, the estimation approach for "termination" (for example, the app described in Example 1) This can be referred to as Roach. In the results shown in Figure 3, the results of the censoring estimation approach The resulting protein identification is related to binding events associated with specific unidentified proteins. This is based on evaluating the occurrence of [unknown]. Therefore, the censoring estimation approach is based on [unknown] When determining the identity of the protein, non-binding outcomes are not considered.

[0152] This type of censoring estimation approach considers all possible combined outcomes. (For example, binding and non-binding outcomes related to specific unidentified proteins) This contrasts with a “non-censored” approach (for both combined measured outcomes). In this embodiment, a particular binding measurement or binding measurement outcome is more prone to error. The expected binding measurement outcome for the protein (for example) For example, the possibility of deviating from the probability that the binding measurement outcome is generated by the protein. A censorship approach may be applicable when such a situation is anticipated. For example, In affinity reagent binding experiments, the binding measurement outcome and the non-binding measurement outcome The probability can be calculated based on binding to denatured proteins that have a mostly linear structure. Under these conditions, the epitope may be easily accessible to the affinity reagent. However, in some embodiments, binding measurement in assayed protein samples These can be collected under non-denaturing or partially denaturing conditions, under which the protein is It exists in a "folded" state with a remarkable three-dimensional structure, which in many cases The affinity reagent binds to an epitope on a protein that is accessible in a linear form, and the structure is folded. In its folded state, it can be made inaccessible due to steric hinderance. For example, with respect to a certain protein, the epitope recognized by the affinity reagent is folded. If it is located in a structurally accessible region of the obtained protein, it is obtained in an unknown sample. Furthermore, experimental binding measurements are calculated from the probability of binding originating from linearized proteins. , and a match can be expected. However, for example, recognized by affinity reagents If the epitope is structurally inaccessible, the binding originates from the linearized protein. It is expected that there are more unconnected outcomes than predicted from the calculated probabilities. Furthermore, based on specific conditions surrounding the protein, the three-dimensional structure can be several It can be formed in various possible three-dimensional configurations, and each of the various possible three-dimensional configurations This relates to binding to a specific affinity reagent based on the degree of accessibility of the desired affinity reagent. They may have unique predictions.

[0153] Therefore, non-binding outcomes deviate from the binding probabilities calculated for each protein. It is expected that this will happen, and only the combined outcome will be considered, presumed censorship An approach may be appropriate. For example, an estimation approach for "censoring" as provided in Figure 3. In this case, only the measured coupled outcomes are considered (in other words, non-coupled outcomes) (either the outcome is not measured, or the measured non-boundary outcome is not considered), Therefore, the M measured binding outcomes yielded binding measurements, which are binding measurements. All N measured coupled outcomes, including both outcome and uncoupled measured outcomes. Although it is a subset of the Tocam, only this subset is considered when determining the probability of the combined outcome set. This can be described by the following expression: P(outcome set | protein) = P(binding event 1 | protein) * P(binding event 1) T2 | Protein) * … * P (Binding event M | Protein)

[0154] When applying a censoring approach, to correct for bias, the scaling factor P(joint) is used. It may be appropriate to apply this to outcome sets (proteins). For example, longer Proteins generally have a higher probability of producing potential binding outcomes. For example, because they contain more potential binding sites. The scaled likelihood SL is used for P (binding outcome set | protein) at the M site. The number of unique combinations of binding sites, which is based on the number of potential binding sites on the protein. It can be produced from protein, and by dividing it, each symptom It can be calculated for coproteins. Proteins of length L that have a trimer recognition site. Regarding this, L-2 potential binding sites may exist (for example, in a complete protein sequence, For each possible subarray of length L, therefore the following holds: TIFF2026076191000009.tif12159

[0155] The probability of any candidate protein being selected from a population of Q possible candidate proteins is Given a set of outcomes, it can be expressed as follows: TIFF2026076191000010.tif13128

[0156] Censored protein estimation approach vs. uncensored protein estimation approach The performance of each embodiment is plotted in Figure 3. The data plotted in Figure 3 is provided in Table 1. .

[0157] (Table 1) TIFF2026076191000011.tif83128

[0158] In the comparison shown in Figure 3, the sensitivity of protein identification (for example, the unique type identified) Estimated censoring (percentage of protein) used in linear protein substrates For both the and uncensored estimates, plot against the number of affinity reagent groups measured. The affinity reagent used targets the topmost and most abundant trimer in the proteome. Furthermore, each affinity reagent exhibits off-target affinity for four additional random trimers. It has properties. When a group of 100 affinity reagents is used, the uncensored approach is 10 The uncensored estimation is more than twice as effective as the censored approach. The degree to which the performance is better than the estimation decreases when a larger number of groups are used.

[0159] Example 4: Protein identification of random false-negative and false-positive affinity reagent binding. tolerance In some cases, there are many false-negative binding outcomes regarding affinity reagent binding. It is possible that false negatives may occur. The combined outcome of "false negatives" occurs less frequently than expected. This manifests as affinity reagent binding measurements. Such "false negative" outcomes include, for example, For example, binding detection method, binding conditions (e.g., temperature, buffer composition, etc.), protein sample This can occur due to degradation or degradation of affinity reagent stocks. Censored proteins. The impact of false negative measurements in identification approaches and uncensored protein identification approaches. To determine the resonance, the subset of affinity reagent measurement groups is 1 out of 10, or 1 out of 100. , one out of 1,000, one out of 10,000, or one out of 100,000, By artificially exchanging observed coupled events for uncoupled events in silico, They intentionally degraded it. Out of the 300 affinity reagents, 0, 1, 50, 100, and 20 were used. Either 0 or 300 samples were degraded in this manner. The results are plotted in Figure 4. As shown, censored protein identification approaches and uncensored proteins Both quality identification approaches allow for this type of random false-negative coupling. Figure 4 shows The data to be lotted is provided in Table 2.

[0160] (Table 2) TIFF2026076191000012.tif241154TIFF2026076191000013.tif145154

[0161] Similarly, tolerance for "false positives" in the binding outcome limits the subset of the binding outcome. This evaluation was performed by swapping non-coupled outcomes with coupled outcomes. The results are provided in Table 3.

[0162] (Table 3) TIFF2026076191000014.tif56154TIFF2026076191000015.tif241154TIFF2026076191000016.tif104154

[0163] These results, plotted in Figure 5, show that as the occurrence of random false positive measurements increases, The performance of the censored protein identification approach is better than that of the uncensored protein identification approach. This also shows a rapid deterioration. However, both approaches show that for each affinity reagent group A false positive rate of 1 in 1000, or 1 in 100 in a subset of affinity reagents. We will allow that proportion.

[0164] Example 5: Protein using overestimated or underestimated affinity reagent binding probabilities Quality estimation performance The sensitivity of protein identification depends on the accurately predicted binding probability of the affinity reagent to the trimer, and Using protein identification that employs overestimated or underestimated binding probabilities, The true joint probability was evaluated as 0.25. The underestimated joint probabilities were 0.05 and 0.1. The probabilities were 0.2, 0.30, 0.50, 0.75, and 0.90. Yes, there were. A total of 300 affinity reagent measurements were obtained. Of the affinity reagents, there were none (0). ), all 300, or subsets (1, 50, 100, 200) were overestimated. or an underestimated binding probability was applied. All else is in protein identification. Therefore, the exact binding probability (0.25) was used. The results of the analysis are provided in Table 4.

[0165] (Table 4) TIFF2026076191000017.tif227170TIFF2026076191000018.tif234170TIFF2026076191000019.tif147170

[0166] These results plotted in Figure 6 suggest that the binding probabilities may not be accurately estimated. In some cases, censored protein identification may be a preferred approach. This indicates.

[0167] Example 6: Protein prediction approach using affinity reagents with unknown binding epitopes performance In some cases, affinity reagents may have several unknown binding sites. A censored protein identification approach and an uncensored protein identification approach using sex reagent binding assays. The sensitivity of the protein identification approach is that it can identify five trimer sites (e.g., one target trimer, and 0.25 (at four random off-target sites) is input into the protein identification algorithm. The comparison was made using affinity reagents that bind with a certain probability to each. A subset of affinity reagents ( 0 out of 300, 1 out of 300, 50 out of 300, 100 out of 300, 20 out of 300 0 or 300 out of 300) will have an additional extra of 1, 4, or 40. It has binding sites, and these sites are 0.05, 0.1, and relative to a random trimer. It had a binding probability of 0.25. The results of the analysis are shown in Table 5.

[0168] (Table 5) TIFF2026076191000020.tif213170TIFF2026076191000021.tif241170TIFF2026076191000022.tif241170

[0169] These results, plotted in Figure 7, show that the uncensored estimates additional hidden join sites. The greater tolerance for inclusion and the performance of both estimation approaches is 30 If 50 out of 0 affinity reagents contain 40 additional binding sites, the affinity is significantly impaired. This indicates that.

[0170] Example 7: Performance of a protein prediction approach using affinity reagents lacking binding epitopes In some cases, there are no annotated binding epi It is inappropriately characterized using tope (e.g., extra expected binding sites). Affinity reagents may exist. In other words, they generate the expected binding probability with respect to affinity reagents. The model used includes extra expected sites that do not exist. Affinity reagent binding. Measurement-based approaches to censored and uncensored protein identification The sensitivity of the approach is determined by random trimer sites (e.g., one target trimer and four non-target sites). (An intentionally created off-target site) is input into the protein identification algorithm with a probability of 0.25. The affinity reagents used for each were compared. A subset of affinity reagents (300 reagents) was used. 0 of them, 1 out of 300, 50 out of 300, 100 out of 300, 200 out of 300, and (or 300 out of 300) will have one, four, or forty extra expected results. It has binding sites, each of which has a binding probability of 0.05 and 0.1 for a random trimer. or has a factor of 0.25 for affinity reagents used by protein estimation algorithms. It was added to the model. The results of the analysis are shown in Table 6.

[0171] (Table 6) TIFF2026076191000023.tif97170TIFF2026076191000024.tif241170TIFF2026076191000025.tif241170TIFF2026076191000026.tif138170

[0172] These results, plotted in Figure 8, show that the uncensored estimates of affinity reagent binding are in line with the model. It has a higher tolerance for including extra expected binding sites, and The performance of both protein identification approaches is such that most affinity reagents are 40 extra expected values. This indicates that the binding site will be damaged to some extent if it is included in the binding site.

[0173] Example 8: Estimation of censoring for affinity reagent binding analysis using an alternative scale transformation strategy The methods described herein can be combined with various probability scaling transformation strategies. Therefore, using affinity reagent binding measurement, the identity of a protein (for example, an unknown protein) can be estimated. This can be applied to the identification of proteins. The censorship estimate approach described in Example 3 Chi is the number of potential binding sites in a protein (protein length - 2), and observed Based on the number of bound outcomes (M), the observed outcomes for the protein Scaling the probability: TIFF2026076191000027.tif12128

[0174] The method described herein is another method for calculating scaled likelihoods. This approach can be applied to the affinity assay used to measure proteins. Model the probability of generating N binding events for a protein of length k from a set of drugs. Another approach to normalization is to transform the probability and then scale it based on that probability. Apply. First, for each probe, the probe identifies the unknown identity in the sample. The probability of binding to the trimer is calculated: TIFF2026076191000028.tif15128 Here, P (trimer) j ) is compared to the total number of all 8,000 trimers in the proteome. This is the frequency of trimer existence. For any protein of length k, probe i is tan The probability of binding to a protein can be expressed as follows: P (Protein binding | Probe) i , k) = 1 - (1 - P(trimer bond | probe i )) k-2

[0175] For a protein of length k, the number of successful binding events observed over n trials is It can also follow a Poisson binomial distribution, where n is the number of probes performed on the protein. The number of coupled measurements and the distribution parameter p プローブ, k The probability of success for each trial Show the rate: p プローブ, k = [P(bond | probe1, k), P(bond | probe2, k), P(bond | probe (3, k) ... P (binding | probe) n, k)]

[0176] Using a specific set of probes, N binding events are generated from a protein of length k. The probability of an event occurring is parameterized by p and evaluated at N, according to the probability of a Poisson binomial distribution. Rate-mass function (PMF) PoiBin ) can be expressed as: P(N binding events | probe, k) = PMF PoiBin (N, p プローブ, k )

[0177] The scaled likelihood of a particular set of outcomes is calculated based on this probability. : TIFF2026076191000029.tif11128

[0178] Example 9: Use of randomly selected affinity reagents The methods described herein may be applied to any set of affinity reagents. For example, protein identification approaches target the most abundant trimers in the proteome. It can be applied to nomination reagents or affinity reagents that target random trimers. A affinity reagent that targets the top and most abundant trimers (300) in the proteome, without Affinity reagents that target 300 intentionally selected trimers, or the least abundant in the proteome. The results from human protein prediction analysis using affinity reagents targeting 300 trimers are as follows: This is shown in Tables 7a to 7c.

[0179] Table 7a~Table 7c (Table 7a) 300 affinity reagents targeting the least abundant trimer in the proteome. TIFF2026076191000030.tif49128

[0180] (Table 7b) 300 affinity reagents targeting random trimers in the proteome. TIFF2026076191000031.tif145128TIFF2026076191000032.tif241122TIFF2026076191000033.tif241122TIFF2026076191000034.tif241122 TIFF2026076191000035.tif241122TIFF2026076191000036.tif241122TIFF2026076191000037.tif241122TIFF2026076191000038.tif241122 TIFF2026076191000039.tif241122TIFF2026076191000040.tif241122TIFF2026076191000041.tif241122TIFF2026076191000042.tif241122 TIFF2026076191000043.tif241122TIFF2026076191000044.tif241122TIFF2026076191000045.tif241122TIFF2026076191000046.tif145128

[0181] (Table 7c) 300 affinity reagents targeting the most abundant trimers in the proteome. TIFF2026076191000047.tif49128

[0182] These results are plotted in Figure 9. In all cases, each affinity reagent is a standard It has a binding probability of 0.25 for the target trimer, and an additional trimer randomly selected It had a binding probability of 0.25 to 4. The performance of each affinity reagent set is such that the sensitivity (even Each affinity reagent set is measured based on the percentage of identified proteins. The evaluation was repeated 5 times, where the performance of each iteration was plotted as a point, and The vertical lines connect repeated measurements from the same set of affinity reagents. The results from the affinity reagent set, which consists of a rich array of 300 affinity reagents, are blue. The lowest 300 are green. Of the 300 affinity reagents targeting random trimers, the total is... 100 different sets were generated and evaluated. Each of those sets was grayed out. The pattern is represented by a set of five gray dots (one representing each repetition) connected by vertical lines. It is determined that, based on the uncensored estimation used in this analysis, a richer trimer is targeted. Targeting improves identification performance compared to targeting random trimers.

[0183] Example 10: Affinity reagent having a biosimilar off-target site The methods described herein involve different types of off-target binding sites (epito This can be applied to affinity reagent binding experiments using affinity reagents having a (-p) property. In this study, the performance of two classes of affinity reagents is compared: random affinity reagent, and Biosimilar affinity reagents. The results from these evaluations are shown in Tables 8a to 8d.

[0184] Table 8a~Table 8d (Table 8a) Biosimilars have off-target sites and are the most in the proteome. The performance of censorship estimation is achieved using affinity reagents that target 300 abundant trimers. TIFF2026076191000048.tif35128

[0185] (Table 8b) Biosimilars have off-target sites and are the most in the proteome. The performance of uncensored estimation using affinity reagents that target a rich array of 300 trimers. TIFF2026076191000049.tif35128

[0186] (Table 8c) It has random off-target sites and is the most abundant in the proteome. Performance of censoring estimation using affinity reagents targeting 300 trimers. TIFF2026076191000050.tif35128

[0187] (Table 8d) It has random off-target sites and is the most abundant in the proteome. Performance of uncensored estimation using affinity reagents targeting 300 trimers. TIFF2026076191000051.tif35128

[0188] Unlike random affinity reagents, biosimilar affinity reagents target epitopes and biochemical They have scientifically similar off-target binding sites. Random affinity reagents and bios Both Miller affinity reagents bind to their target epitopes (e.g., trimers) with a certain probability. Recognized at 0.25. Each of the randomly selected affinity reagents has a binding probability of 0.25, randomly selected. It has four selected off-target trimer binding sites. In contrast, "biosimilars" The four off-target binding sites of the affinity reagent are most similar to the target trimer of the affinity reagent. These are four trimers, which bind with a probability of 0.25. Their biosimilar affinity Regarding the reagents, the similarity between trimer sequences is BL for each amino acid pair at each sequence position. It is calculated by summing the OSUM62 coefficients. A random set of affinity reagents and Both sets of iosimilar affinity reagents are among the most abundant and top-tier in the human proteome. The target is 300 mers, where the degree of abundance indicates that one or more examples of trimers are included. It is measured as the number of unique proteins. Figure 10 shows random off-target sites. Affinity reagents (blue) or biosimilar affinity reagents having off-target sites (Orange) Percentage of proteins identified in human samples when used Regarding this, the censored protein estimation approach (dashed line) and the uncensored protein estimation approach The performance of the estimation approach (solid line) is shown.

[0189] In this comparison, the uncensored estimation performed better than the censored estimation. The estimation of censoring was performed better in the case of biosimilar affinity reagents, and Censoring estimation is performed more effectively when random affinity reagents are used.

[0190] Alternatively, one could use affinity reagents that target the most abundant trimer in the proteome. Rather, the optimal set of trimer targets is the candidate protein that can be measured (e.g., human protein (Ohm), the type of protein estimation performed (censored or uncensored), and the method used. Based on the type of affinity reagent used (random or biosimilar), a specific app Regarding Roach, a choice can be made. The "greedy" algorithm, as described later, has the optimal affinity. To select a set of reagents, the following may be used: 1) Initialize an empty list of affinity reagents (ARs) to be selected. 2) A set of candidate ARs (for example, each having a random off-target site) Initialize a collection of 8,000 ARs (each targeting a unique trimer). 3) For (for example, all human proteins in the Uniprot reference proteome) To optimize this, a set of protein sequences is selected. 4) Repeat the following until the desired number of ARs are selected: a. Regarding each candidate AR: i. Simulate the binding of candidate ARs to the protein set. ii. Simulated coupling measurements from candidate ARs, and all previously selected A Using binding measurements simulated from R, protein identification was performed for each protein. To administer. iii. The exact protein for each protein, determined by protein estimation. A score is calculated for each candidate AR by summing the probabilities of substance identification. b. Add the AR with the best score to the set of selected ARs, and then make it a candidate AR. Remove from the list.

[0191] The greedy approach targets the top 4,000 most abundant trimers in the human proteome. This is either a random population of affinity reagents or a population of biosimilar affinity reagents. Then, to select the 300 most suitable affinity reagents, optimization was performed, but the process was terminated. This was performed for both protein estimation and uncensored protein estimation. The results from the chemical analysis are provided in Tables 9a to 9d.

[0192] Table 9a~Table 9d (Table 9a) Biosimilars have off-target sites and are the most important in the proteome. Performance of censorship estimation using affinity reagents targeting 300 appropriate trimers. TIFF2026076191000052.tif35128

[0193] (Table 9b) Biosimilars have off-target sites and are the most important in the proteome. Performance of uncensored estimation using affinity reagents targeting 300 suitable trimers. TIFF2026076191000053.tif35128

[0194] (Table 9c) It has random off-target sites and is the optimal three in the proteome. Performance of censorship estimation using affinity reagents targeting 300 mers. TIFF2026076191000054.tif35128

[0195] (Table 9d) It has random off-target sites and is the optimal three in the proteome. Performance of uncensored estimation using affinity reagents targeting 300 mers. TIFF2026076191000055.tif35128

[0196] For both censored and uncensored protein estimation, optimal The performance of the modified probe set is plotted in Figure 11.

[0197] The use of the set of affinity reagents selected by the greedy optimization algorithm is terminated. We use both censored and uncensored protein estimation approaches. The performance of both random affinity reagent sets and biosimilar affinity reagent sets is being evaluated. , improve. In addition, a random set of affinity reagents allows the greedy approach to evaluate affinity reagents When used for selection, it behaves much like a set of biosimilar affinity reagents. do.

[0198] Example 11: Protein estimation using a mixture of affinity reagents The method described herein involves measuring proteins using a mixture of affinity reagents. It can be applied to analyze and / or identify the quality by a mixture of affinity reagents. When assayed, the probability that a particular protein produces a binding outcome is as follows: It can be calculated as follows: 1) The average probability of nonspecific epitope binding for each affinity reagent in the mixture. Calculate TIFF2026076191000056.tif4128. 2) The number of binding sites on a protein is determined by the length (L) of the protein and the affinity reagent. Calculated based on epitope length (K): Number of binding sites = L - K + 1. Nonspecific binding. The probability that the event does not occur is, The filename is TIFF2026076191000057.tif5128. 3) The probability that no epitope-specific binding event occurs for each affinity reagent in the mixture. The calculation is as follows: (TIFF2026076191000058.tif191284) Regarding proteins, the probability that a mixture produces a non-binding outcome is as follows: (TIFF2026076191000059.tif111285) The probability that a mixture produces a binding outcome is as follows: P(bound | protein) = 1 - P(unbound | protein)

[0199] To calculate the probability of binding or non-binding outcomes from a protein mixture. This approach analyzes the performance of a mixture of affinity reagents for protein identification. Therefore, it was used in combination with the methods described herein. Each affinity reagent binds to its target trimer epitope with a probability of 0.25, and the epi It binds to the four trimers most similar to the tope target with a probability of 0.25. Regarding the similarity of trimers, the amino acids at each sequence position in the trimers being compared are... It is calculated by summing the coefficients from the BLOSUM62 substitution matrix. In addition, each Affinity reagents were calculated using the BLOSUM62 substitution matrix, and are found at off-target sites and target sites. According to the sequence similarity with respect to the trimer, 20 additional join probabilities are assigned with scaled-up join probabilities. It binds to off-target sites. The probability for these additional off-target sites is: The following applies: TIFF2026076191000060.tif5128 Here S OT This is the BLOSUM62 similarity between the off-target site and the target site, and S self This represents the BLOSUM62 similarity between the target sequence and itself. 2.45 x 10 8 Joint probabilities below All of the off-target sites have a binding probability of 2.45 x 10⁻⁶. 8 It is adjusted to have. The nonspecific epitope binding probability in this example is 2.45 x 10⁻⁶. 8 That is the case.

[0200] The optimal set of 300 affinity reagent mixtures is suitable for both censored and uncensored applications. The following were generated using a greedy approach for protein identification: 1) Initialize an empty list of affinity reagent (AR) mixtures to be selected. 2) Candidate affinity reagent (in this example, the greedy approach described in detail in Example 10) Initialize a list of 300 optimal options calculated using the formula. . 3) For (for example, all human proteins in the Uniprot reference proteome) To optimize this, a set of protein sequences is selected. 4) Repeat the following until the desired number of AR mixtures are produced: a. Initialize an empty mixture. b. Regarding each candidate AR: i. Using the current mixture with added candidate ARs, the binding outcome was determined To simulate. ii. Simulated binding measurements from i., and from the previously generated mixture Protein identification is performed for each protein using muted binding measurements. iii. The exact protein for each protein, determined by protein estimation. By summing the probabilities of substance identification, a score is calculated for mixtures that have this candidate AR. To determine. c. Add the candidate AR with the highest score to the mixture. d. With respect to each candidate AR that was not previously present in the mixture, i The candidates are scored as in ~iii, and the candidate with the highest score is the mixture. If it has a higher score than the previously added candidate, add it to the mixture, and Then, this process is repeated. The mixture is then mixed with the candidate AR that received the highest score, which was previously added. When the score of the mixture decreases compared to the candidate, or when all candidate ARs are in the mixture It will be completed once it is added.

[0201] Figure 12 shows the results of unmixed candidate affinity reagents and mixed reagents for censored protein estimation. This indicates the sensitivity of protein identification when used in conjunction with uncensored protein estimation. The data plotted in Figure 12 are shown in Tables 10a to 10b.

[0202] Table 10a~Table 10b (Table 10a) The binding of individual probes (unmixed) or mixtures of probes (mixed) The performance of censorship estimation using measurements taken in [location] TIFF2026076191000061.tif56141

[0203] (Table 10b) The binding of individual probes (unmixed) or mixtures of probes (mixed) Performance of uncensored estimation using measurements taken in TIFF2026076191000062.tif56142

[0204] Using a mixture improves performance when an uncensored estimate is used, but when censored... When estimations are used, they can have an adverse effect on performance.

[0205] Example 12 - Glycan identification using a database of 7 candidate glycans Let's consider the situation where the database contains seven candidate glycans: TIFF2026076191000063.tif61162

[0206] In addition, the experiment found four affinity groups, each with a 25% likelihood of binding to a given disaccharide. The procedure is carried out using reagents (AR). Other disaccharides to which these reagents bind are listed in the database. It is not found in any glycans.

[0207] A hit table is constructed for each affinity reagent in the database. (Rows = Affinity reagents #1~#4, Columns = SEQ IDs) TIFF2026076191000064.tif38170

[0208] In particular, this information arrives gradually, and therefore can be calculated iteratively. As shown, from the hit table, P(glycan_i | AR_j) gives the probability matrix It is evaluated in order to generate. For a given entry, the hit table >= 1. Therefore, P_landing_AR_n = true landing rate = 0.25 is used; instead, hittable = 0 In that case, please note that P (detector error) = 0.00001 is used. TIFF2026076191000065.tif67166

[0209] Note that many cells contain a probability of 0.00001. This small probability is due to the detector An error may be the cause. First, the probability of glycans is not normalized, and each symptom The proglycans are calculated as a product of probabilities: TIFF2026076191000066.tif16153

[0210] Next, the size normalization is calculated, which means that several affinity reagents are given a certain amount of glycans. The number of possible configurations for landing on the can is expressed as a function of the number of potential binding sites on the glycan. The normalization of 's' is expressed as the word Choose(part_i, n). For example, candidate ID 52 has 6 It has a disaccharide moiety and a normalization of size [6 choose 4] which is 15. If there are more binding events than there are disaccharide sites, the size normalization factor is set to 1. The probability of each glycan, which is not normalized, takes this size correction into account. It is normalized by dividing by the normalized size, as shown below: TIFF2026076191000067.tif21165

[0211] Next, the probability sets across the entire database so that they sum up to 1. This is normalized. This is done by summing the values ​​so that the probability of the size being normalized is 0.00390641, and so In order to achieve the final balanced probability, this normalization normalizes the size. This is achieved by dividing each of the probabilities: TIFF2026076191000068.tif23170

[0212] item 1. To repeatedly identify candidate proteins in a sample of an unknown protein, A method performed by a tar, comprising the following steps: (a) Each of the multiple affinity reagent probes for the unknown protein in the sample The process involves receiving information about the coupling measurement by the computer, Each of the compatible reagent probes is one or more candidate proteins from among multiple candidate proteins. A process designed to selectively bind to a substance; (b) At least a portion of the information from the binding measurement is data including multiple protein sequences. A step of comparing each protein sequence with the base using the computer, However, a step corresponding to one candidate protein among the plurality of candidate proteins; and (c) With respect to one or more of the candidate proteins among the plurality of candidate proteins Then, the probability that each of the one or more candidate proteins is present in the sample is, The information on the binding measurement of each of the candidate proteins is at least one Based on the comparison of the portion with the database containing the plurality of protein sequences , a process of repeatedly generating by the computer. 2. The process of generating the multiple probabilities is, Additional information on binding measurements for each of the multiple additional affinity reagent probes, repeated to receive The further comprising additional affinity reagent probes, each of which is among the multiple candidate proteins The following is described in item 1, which is configured to selectively bind to one or more candidate proteins. method. 3. For each of the one or more candidate proteins, if the candidate protein is A step of generating a confidence level consistent with one of the unknown proteins in the sample. The method described in item 1, further including the method described in item 1. 4. The process of generating the aforementioned probability is The error rate of the detector, related to the information from the aforementioned coupled measurement, should be taken into consideration. The method described in item 1, including the method described in item 1. 5. The error rate of the detector is used to obtain information on the coupled measurement. The method described in item 4, obtained from the specifications of multiple detectors. 6. The method described in item 4, wherein the error rate of the detector is set to the estimated error rate of the detector. 7. The error rate of the detector is set by the user of the computer. The method described in item 6. 8. The method described in item 6, wherein the estimated error rate of the detector is approximately 0.001. 9. The process of repeatedly generating the aforementioned multiple probabilities is: From the subsequent iterations, one or more candidate proteins are removed from the plurality of candidate proteins. To leave It further includes, thereby, how many are necessary to carry out the repeated generation of the probabilities. The method described in item 1 to reduce the repetition. 10. Removing one or more candidate proteins is related to the candidate proteins. The method according to item 9, which is based at least on a predetermined standard for the binding measurement. 11. The aforementioned prescribed standards are Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate proteins having the above characteristics The method described in item 10, including the method described in item 10. 12. Each of the above probabilities is normalized with respect to the length of the candidate protein, as described in item 1. Method of loading. 13. Each of the above probabilities is normalized with respect to the sum of the probabilities of the multiple candidate proteins. The method described in item 1. 14. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, as described in item 1. Method of loading. 15. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 1 Method of description. 16. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 1 Method of description. 17. The plurality of affinity reagent probes include more than 500 affinity reagent probes, as described in item 1. Method of loading. 18. The method described in item 1, wherein the aforementioned probability is repeatedly generated until a predetermined condition is met. 19. The predetermined conditions generate each of the multiple probabilities with at least 90% confidence. The method described in item 18, including the method described in item 18. 20. The predetermined conditions generate each of the plurality of probabilities with at least 95% confidence. The method described in item 19, including the above. 21. The predetermined conditions generate each of the plurality of probabilities with at least 99% confidence. The method described in item 20, including the above. 22. A paper or electronic report identifying one or more unknown proteins in the sample. The process of generating a t The method described in item 1, further including the method described in item 1. 23. The method described in item 1, wherein the sample includes a biological sample. 24. The method described in item 23, wherein the biological sample is obtained from the subject. 25. An engineering method for identifying the disease state in the subject based on at least the aforementioned multiple probabilities. degree The method described in item 24, further including the method described in item 24. 26. Computers to identify candidate proteins in samples of unknown proteins. A method carried out by the following steps: (a) Each of the multiple affinity reagent probes for the unknown protein in the sample The process involves receiving information about the coupling measurement by the computer, Each of the compatible reagent probes is one or more candidate proteins from among multiple candidate proteins. A process designed to selectively bind to a substance; (b) At least a portion of the information from the binding measurement is data including multiple protein sequences. A step of comparing each protein sequence with the base using the computer, However, a step corresponding to one candidate protein among the plurality of candidate proteins; and (c) At least a portion of the information of the binding measurement, including the plurality of protein sequences Based at least on the comparison with the database, the plurality of candidate proteins A process to remove one or more candidate proteins. 27. The step of removing one or more candidate proteins is related to the candidate proteins. The method according to item 26, which is based at least on a predetermined standard for the coupling measurement. 28. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate proteins having the above characteristics The method described in item 27, including the method described in item 27. 29. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 26 Method of description. 30. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 26 Method of description. 31. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 26 Method of description. 32. The plurality of affinity reagent probes include more than 500 affinity reagent probes, as described in item 26. Method of loading. 33. A paper or electronic report identifying one or more unknown proteins in the sample. The process of generating a t The method described in item 26, further including the method described in item 26. 34. The method described in item 26, wherein the sample includes a biological sample. 35. The method described in item 34, wherein the biological sample is obtained from the subject. 36. Based at least on the identified candidate protein, the disease state in the subject Steps to identify the state The method described in item 35, further including the method described in item 35. 37. Computer-aided systems for iteratively identifying candidate glycans in samples of unknown glycans. A method carried out by - and comprising the following steps: (a) Each of the multiple affinity reagent probes for the unknown glycan in the sample A step of receiving the binding measurement for the affinity reagent by the computer, Each probe is selective for one or more candidate glycans among multiple candidate glycans. A process that is set up to be combined; (b) The binding measurement is performed on a database containing multiple glycan sequences using the computer. A step of comparing by a ter, wherein each glycan sequence is compared among the plurality of candidate glycans. A process corresponding to one candidate glycan; and (c) With respect to each of the one or more candidate glycans among the plurality of candidate glycans, The probability that each of the one or more candidate glycans is present in the sample is the binding For measurement, a plurality of glycans corresponding to one of the plurality of candidate glycans. Based on the comparison with the database containing the Lycan sequence, the computer Therefore, it is a process of repeated generation. 38. The process of generating the aforementioned multiple probabilities is Additional information on binding measurements for each of the multiple additional affinity reagent probes, repeated to receive The further comprises, each of which additional affinity reagent probes is one of the multiple candidate glycans. The method described in item 37, which is configured to selectively bind to one or more candidate glycans. Law. 39. For each of the one or more candidate glycans, if the candidate glycan is the same as the trial A process of generating a confidence level consistent with one of the unknown glycans in the material. The method described in item 37, further including the method described in item 37. 40. The process of generating the aforementioned probability is The error rate of the detector, related to the information from the aforementioned coupled measurement, should be taken into consideration. The method described in item 37, including the method described in item 37. 41. The error rate of the detector is used to obtain information on the coupled measurement. The method described in item 40, obtained from the specifications of multiple detectors. 42. The method according to item 40, wherein the error rate of the detector is set to the estimated error rate of the detector. 43. The estimated error rate of the detector is set by the user of the computer. , the method described in item 42. 44. The method described in item 42, wherein the estimated error rate of the detector is approximately 0.001. 45. The process of repeatedly generating the above-mentioned multiple probabilities is: From the subsequent iterations, one or more candidate glycans are removed from the plurality of candidate glycans. to It further includes, thereby, how many are necessary to carry out the repeated generation of the probabilities. The method described in item 37 to reduce that repetition. 46. ​​The step of removing one or more candidate glycans is related to the candidate glycans The method described in item 45, which is based at least on a predetermined standard for the binding measurement. 47. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate glycans having The method described in item 46, including the method described in item 46. 48. Each of the above probabilities for several potential binding sites of the candidate glycan The method used for normalization is described in item 37. 49. Each of the above probabilities is normalized with respect to the sum of the probabilities of the multiple candidate glycans. The method described in item 37. 50. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 37 Method of description. 51. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 37 Method of description. 52. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 37 Method of description. 53. The plurality of affinity reagent probes include more than 500 affinity reagent probes, as described in item 37. Method of loading. 54. The method described in item 37, wherein the probability is generated repeatedly until a predetermined condition is met. . 55. The predetermined conditions generate each of the multiple probabilities with at least 90% confidence. The method described in item 54, including the method described in item 54. 56. The predetermined conditions generate each of the plurality of probabilities with at least 95% confidence. The method described in item 55, including the above. 57. The predetermined conditions generate each of the plurality of probabilities with at least 99.999% confidence. The method described in item 56, including doing so. 58. A paper or electronic report identifying one or more unknown glycans in the sample. The process of generating a The method described in item 37, further including the method described in item 37. 59. The method described in item 37, wherein the sample includes a biological sample. 60. The method described in item 59, wherein the biological sample is obtained from the subject. 61. An engineering method for identifying the disease state in the subject based on at least the above-mentioned multiple probabilities. degree The method described in item 60, further including the method described in item 60. 62. Computer-aided identification of candidate glycans in a sample of an unknown glycan. A method that is carried out by the following steps: (a) Each of the multiple affinity reagent probes for the unknown glycan in the sample A step of receiving the binding measurement for the affinity reagent by the computer, Each probe is selective for one or more candidate glycans among multiple candidate glycans. A process that is set up to be combined; (b) At least a portion of the binding measurement is performed in a database containing multiple glycan sequences. In contrast, the process of comparison by the computer, wherein each glycan sequence is the same as the multiple A process corresponding to one candidate glycan among a number of candidate glycans; and (c) At least a portion of the information of the binding measurement includes the plurality of glycan sequences Based at least on the comparison against the database, from the plurality of candidate glycans A step to remove one or more candidate glycans. 63. The step of removing one or more candidate glycans is related to the candidate glycans. The method according to item 62, which is based at least on a predetermined standard for the binding measurement. 64. The aforementioned prescribed standards are Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate glycans having The method described in item 63, including the method described in item 63. 65. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 62 Method of description. 66. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 62 Method of description. 67. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 62 Method of description. 68. The plurality of affinity reagent probes include more than 500 affinity reagent probes, as described in item 62. Method of loading. 69. A paper or electronic report identifying one or more unknown glycans in the sample. The process of generating a The method described in item 62, further including the method described in item 62. 70. The method described in item 62, wherein the sample includes a biological sample. 71. The method described in item 70, wherein the biological sample is obtained from the subject. 72. Based at least on the identified candidate glycans, the disease status in the subject. The process of identifying The method described in item 71, further including the method described in item 71. 73. Any of the claims wherein the binding measurement includes measuring the binding of an affinity reagent to a glycan. The method described in item 1. 74. Any of the claims wherein the binding measurement includes measuring the non-binding of an affinity reagent to a glycan. The method described in item 1. 75. The predetermined conditions are such that the multiple probabilities are met with at least 99.9999999999999% confidence. The method described in item 57, including generating them. 76. The predetermined conditions are such that the probability of the above multiple probabilities is at least 99.99999999999999% of the probability of the above multiple probabilities being met. The method described in item 57, including generating them. 77. The predetermined conditions are such that the multiple probabilities are met with at least 99.999999999999999% confidence. The method described in item 57, including generating each of them. 78. Computers for repeatedly identifying candidate metabolites of unknown metabolites in a sample. A method carried out by the following steps: (a) Each of the multiple affinity reagent probes for the unknown metabolite in the sample The process involves receiving the binding measurement results by the computer, wherein the affinity reagent Each lobe selectively binds to one or more candidate metabolites from among multiple candidate metabolites. A process that is set up to do so; (b) The binding measurement is performed against a database containing multiple metabolite structures using the computer. - A process of comparison, wherein each metabolite structure is one of the plurality of candidate metabolites Processes corresponding to candidate metabolites; and (c) With respect to one or more candidate metabolites among the plurality of candidate metabolites, The probability that each of one or more candidate metabolites is present in the sample is determined by the binding measurement. , a plurality of metabolite structures corresponding to one candidate metabolite among the plurality of candidate metabolites Based on the comparison with the database, the computer repeats The process of generating. 79. The process of generating the aforementioned multiple probabilities is Additional information on binding measurements for each of the multiple additional affinity reagent probes, repeated to receive The further comprising, each additional affinity reagent probe is one of the multiple candidate metabolites Alternatively, the method described in item 78, which is configured to selectively bind to multiple candidate metabolites. 80. With respect to each of one or more candidate metabolites, if the candidate metabolite is present in the sample A process to generate a confidence level consistent with one of the unknown metabolites. The method described in item 78, further including the method described in item 78. 81. The process of generating the aforementioned probability is The error rate of the detector, related to the information from the aforementioned coupled measurement, should be taken into consideration. The method described in item 78, including the method described in item 78. 82. The error rate of the detector is used to obtain information on the coupled measurement. The method described in item 81, obtained from the specifications of multiple detectors. 83. The method described in item 81, wherein the error rate of the detector is set to the estimated error rate of the detector. 84. The estimated error rate of the detector is set by the user of the computer. , the method described in item 83. 85. The method described in item 83, wherein the estimated error rate of the detector is approximately 0.001. 86. The process of repeatedly generating the aforementioned multiple probabilities is: From the subsequent iterations, one or more candidate metabolites are removed from the plurality of candidate metabolites. and It further includes, thereby, how many are necessary to carry out the repeated generation of the probabilities. The method described in item 78 to reduce that repetition. 87. Removing one or more candidate metabolites is related to the candidate metabolites The method described in item 86, which is based at least on a prescribed standard for binding measurement. 88. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate metabolites having The method described in item 87, including the method described in item 87. 89. Each of the above probabilities is positive for several potential binding sites of the candidate metabolite. The method described in item 78 will be standardized. 90. Each of the above probabilities is normalized with respect to the sum of the probabilities of the multiple candidate metabolites. , the method described in item 78. 91. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 78 Method of description. 92. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 78 Method of description. 93. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 78 Method of description. 94. The plurality of affinity reagent probes include more than 500 affinity reagent probes, as described in item 78. Method of loading. 95. The method described in item 78, wherein the probability is generated repeatedly until a predetermined condition is met. . 96. The predetermined conditions generate each of the multiple probabilities with at least 90% confidence. The method described in item 95, including the method described in item 95. 97. The predetermined conditions generate each of the plurality of probabilities with at least 95% confidence. The method described in item 96, including the above. 98. The predetermined conditions generate each of the plurality of probabilities with at least 99.999% confidence. The method described in item 97, including doing so. 99. A paper or electronic report identifying one or more unknown metabolites in the sample. The process of generating The method described in item 78, further including the method described in item 78. 100. The method according to item 78, wherein the sample includes a biological sample. 101. The method described in item 100 for obtaining the biological sample from the subject. 102. An engineering method for identifying the disease state in the subject based on at least the above-mentioned multiple probabilities. degree The method described in item 101, further including the method described in item 101. 103. Computer-assisted identification of candidate metabolites in a sample of unknown metabolites. A method that is carried out, comprising the following steps: (a) Each of the multiple affinity reagent probes for the unknown metabolite in the sample The process involves receiving the binding measurement results by the computer, wherein the affinity reagent Each lobe selectively binds to one or more candidate metabolites from among multiple candidate metabolites. A process that is set up to do so; (b) At least a portion of the binding measurement is compared with a database containing multiple metabolite structures. The process involves comparing the metabolite structures using the computer, wherein each metabolite structure is one of the multiple A process corresponding to one of the candidate metabolites; and (c) At least a portion of the information of the binding measurement, including the plurality of metabolite structures Based at least on the comparison against the database, one of the multiple candidate metabolites Alternatively, a process to remove multiple candidate metabolites. 104. The step of removing one or more candidate metabolites is related to the preceding The method described in item 103, which is based at least on the prescribed criteria for the measurement of the combination. 105. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate metabolites having The method described in item 104, including the method described in item 104. 106. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 10 Method 3. 107. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 1 Method as described in 03. 108. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 1 Method as described in 03. 109. The plurality of affinity reagent probes include more than 500 affinity reagent probes, item 103 Method of description. 110. A paper or electronic report identifying one or more unknown metabolites in the sample. The process of generating a The method described in item 103, further including the method described in item 103. 111. The method according to item 103, wherein the sample includes a biological sample. 112. The method described in item 111, wherein the biological sample is obtained from the subject. 113. Based at least on the identified candidate metabolites, the disease state in the subject is determined Identification process The method described in item 112, further including the method described in item 112. 114. The binding measurement includes the measurement of the binding of affinity reagents to metabolites, as described in any of the above items. Method of loading. 115. The binding measurement includes any of the above items, including the measurement of the non-binding of affinity reagents to metabolites. Method of description. 116. The predetermined conditions are such that each of the plurality of probabilities is met with at least 99.99999% confidence. The method described in item 98, including generating. 117. The predetermined conditions are such that each of the plurality of probabilities is met with at least 99.999999% confidence. The method described in item 98, including generating. 118. The predetermined conditions are determined to be at least 99.9999999% reliable in each of the above multiple probabilities. The method described in item 98, which includes generating. 119. The predetermined conditions are determined to be at least 99.99999999% reliable in each of the above multiple probabilities. The method described in item 98, which includes generating. 120. The predetermined conditions are determined to be at least 99.99999999% reliable in each of the above multiple probabilities. The method described in item 98, which includes generating. 121. The predetermined conditions are such that each of the above multiple probabilities is determined with at least 99.999999999% confidence. The method described in item 98, which includes generating this. 122. The predetermined conditions are such that each of the above multiple probabilities is determined with at least 99.9999999999% confidence. The method described in item 98, which includes generating this. 123. The predetermined conditions are such that the multiple probabilities are met with at least 99.99999999999% confidence. The method described in item 98, including generating them. 124. The predetermined conditions are such that the multiple probabilities are met with at least 99.9999999999999% confidence. The method described in item 98, including generating them. 125. The predetermined conditions are such that the multiple probabilities are met with at least 99.99999999999999% confidence. The method described in item 98, including generating each of them. 126. The predetermined conditions are such that the multiple probabilities are met with at least 99.999999999999999% confidence. The method described in item 98, including generating each of them. 127. Computer-aided methods for iteratively identifying candidate glycans in samples of unknown glycans. A method carried out by - and comprising the following steps: (a) Each of the multiple affinity reagent probes for the unknown glycan in the sample A step of receiving the binding measurement for the affinity reagent by the computer, Each probe is selective for one or more candidate glycans among multiple candidate glycans. A process that is set up to be combined; (b) The binding measurement is performed on a database containing multiple glycan structures using the computer. A process of comparison by a ter, wherein each glycan structure is compared among the plurality of candidate glycans. A process corresponding to one candidate glycan; and (c) With respect to each of the one or more candidate glycans among the plurality of candidate glycans, The probability that each of the one or more candidate glycans is present in the sample is the binding For measurement, a plurality of glycans corresponding to one of the plurality of candidate glycans. Based on the comparison with the database which includes the Lycan structure, the computer Therefore, it is a process of repeated generation. 128. The process of generating the multiple probabilities is Additional information on binding measurements for each of the multiple additional affinity reagent probes, repeated to receive The further comprises, each of which additional affinity reagent probes is one of the multiple candidate glycans. The following item 127 is configured to selectively bind to one or more candidate glycans. method. 129. With respect to each of the one or more candidate glycans, the candidate glycan is A step of generating a confidence level consistent with one of the unknown glycans in the sample. The method described in item 127, further including the method described in item 127. 130. The process of generating the aforementioned probability is The error rate of the detector, related to the information from the aforementioned coupled measurement, should be taken into consideration. The method described in item 127, including the method described in item 127. 131. The error rate of the detector is used to obtain information on the coupled measurement. This is the method described in item 130, obtained from the specifications of multiple detectors. 132. The method described in item 130, wherein the error rate of the detector is set to the estimated error rate of the detector. . 133. The estimated error rate of the detector is set by the user of the computer. , the method described in item 132. 134. The method according to item 132, wherein the estimated error rate of the detector is approximately 0.001. 135. The process of repeatedly generating the aforementioned multiple probabilities is: From the subsequent iterations, one or more candidate glycans are removed from the plurality of candidate glycans. to It further includes, thereby, how many are necessary to carry out the repeated generation of the probabilities. The method described in item 127 to reduce the repetition. 136. Removing one or more candidate glycans is related to the candidate glycans. The method described in item 135, which is based at least on a predetermined standard for the bonding measurement. 137. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate glycans having The method described in item 136, including the method described in item 136. 138. Each of the above probabilities for several potential binding sites of the candidate glycan The method used for normalization is described in item 127. 139. Each of the above probabilities is normalized with respect to the sum of the probabilities of the multiple candidate glycans. The method described in item 127. 140. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 12 The method described in 7. 141. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 1 The method described in section 27. 142. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 1 The method described in section 27. 143. The plurality of affinity reagent probes include more than 500 affinity reagent probes, item 127 Method of description. 144. The method described in item 127, wherein the aforementioned probability is repeatedly generated until a predetermined condition is met. Law. 145. The predetermined conditions generate each of the multiple probabilities with at least 90% confidence. The method described in item 144, including the above. 146. The predetermined conditions generate each of the plurality of probabilities with at least 95% confidence. The method described in item 145, including the following. 147. The predetermined conditions produce each of the above multiple probabilities with at least 99.999% confidence. The method described in item 146, including the act of accomplishing. 148. Paper or electronic report identifying one or more unknown glycans in the sample. The process of generating a t The method described in item 127, further including the method described in item 127. 149. The method described in item 127, wherein the sample includes a biological sample. 150. The method described in item 149, wherein the biological sample is obtained from the subject. 151. An engineering method for identifying the disease state in the subject based on at least the above-mentioned multiple probabilities. degree The method described in item 150, which further includes the method described in item 150. 152. Computer-aided identification of candidate glycans in a sample of an unknown glycan. A method that is carried out by the following steps: (a) Each of the multiple affinity reagent probes for the unknown glycan in the sample A step of receiving the binding measurement for the affinity reagent by the computer, Each probe is selective for one or more candidate glycans among multiple candidate glycans. A process that is set up to be combined; (b) At least a portion of the binding measurement is performed in a database containing multiple glycan structures In contrast, the process of comparison by the computer, in which each glycan structure is the same as the multiple A process corresponding to one candidate glycan among a number of candidate glycans; and (c) At least a portion of the information of the binding measurement, including the plurality of glycan structures Based at least on the comparison against the database, from the plurality of candidate glycans A step to remove one or more candidate glycans. 153. The step of removing one or more candidate glycans is related to the candidate glycans. The method described in item 152, which is based at least on a predetermined standard for the bonding measurement. 154. The aforementioned prescribed standards, Binding measurement below a predetermined threshold for a first group of affinity reagent probes The one or more candidate glycans having The method described in item 153, including the method described in item 153. 155. The plurality of affinity reagent probes include 50 or fewer affinity reagent probes, item 15 Method described in 2. 156. The plurality of affinity reagent probes include 100 or fewer affinity reagent probes, item 1 The method described in section 52. 157. The plurality of affinity reagent probes include 500 or fewer affinity reagent probes, item 1 The method described in section 52. 158. The plurality of affinity reagent probes include more than 500 affinity reagent probes, item 152 Method of description. 159. Paper or electronic report identifying one or more unknown glycans in the sample. The process of generating a t The method described in item 152, further including the method described in item 152. 160. The method described in item 152, wherein the sample includes a biological sample. 161. The method described in item 160, wherein the biological sample is obtained from the subject. 162. Based at least on the identified candidate glycans, the disease status in the subject. The process of identifying The method described in item 161, further including the method described in item 161. 163. The binding measurement includes any of the above items, including the measurement of the binding of affinity reagents to glycans. Method of description. 164. Any of the above items, in which the binding measurement includes the measurement of the non-binding of affinity reagents to glycans. Methods used. 165. The predetermined conditions are such that each of the plurality of probabilities is met with at least 99.99999% confidence. The method described in item 147, including generating. 166. The predetermined conditions are such that each of the plurality of probabilities is met with at least 99.999999% confidence. The method described in item 147, including generating. 167. The predetermined conditions are determined to be at least 99.9999999% reliable in each of the above multiple probabilities. The method described in item 147, which includes generating. 168. The predetermined conditions are determined to be at least 99.99999999% reliable in each of the above multiple probabilities. The method described in item 147, which includes generating. 169. The predetermined conditions are determined to be at least 99.99999999% reliable in each of the above multiple probabilities. The method described in item 147, which includes generating. 170. The predetermined conditions are such that each of the above multiple probabilities is determined with at least 99.999999999% confidence. The method described in item 147, which includes generating this. 171. The predetermined conditions are such that each of the above multiple probabilities is determined with at least 99.9999999999% confidence. The method described in item 147, which includes generating this. 172. The predetermined conditions are such that the multiple probabilities are met with at least 99.99999999999% confidence. The method described in item 147, including generating them. 173. The predetermined conditions are such that the multiple probabilities are met with at least 99.9999999999999% confidence. The method described in item 147, including generating them. 174. The predetermined conditions are such that the multiple probabilities are met with at least 99.99999999999999% confidence. The method described in item 147, which includes generating each of them. 175. The predetermined conditions are such that the multiple probabilities are met with at least 99.999999999999999% confidence. The method described in item 147, which includes generating each of them.

[0213] Preferred embodiments of the present invention are shown and described herein, but such embodiments It will be obvious to those skilled in the art that the example shown is provided merely as an illustration. This document is not intended to be limited by the specific examples provided herein. The invention is described with reference to the above specification, but the description of embodiments and figures in this specification The term "surface" should not be interpreted in a restrictive sense. Countless changes, modifications, and substitutions are possible. Without departing from the present invention, it will now be conceivable to those skilled in the art. Furthermore, all aspects of the present invention will be recalled. The specific aspects described herein depend on a variety of conditions and variables. It should be understood that this is not limited to fixed descriptions, structures, or relative proportions. Various alternatives to the embodiments of the present invention described in the specification may be adopted when putting the present invention into practice. It should be understood that it can be used in any such way. Therefore, the present invention also applies to any such The intention is to include alternatives, modified versions, altered versions, or equivalents. The attached claims define the scope of the present invention and the scope of these claims. To define the methods and configurations, and how their equivalents are encompassed therein. , intend.

Claims

1. A method for characterizing a protein, comprising the following steps: (a) A step of receiving a binding measurement indicating the binding outcome of a set of affinity reagent probes applied to a protein in a sample, wherein each of the proteins is immobilized at a specific spatial position on a substrate, and the set of affinity reagents comprises a plurality of individual affinity reagents, each of which has known binding properties to different epitopes, including post-translational modifications (PTMs); (b) A step of receiving the binding probability of each of the individual affinity reagent probes to different epitopes including the PTM; and (c) A step of characterizing the modification morphology of the protein in the sample by identifying the PTM in the epitope for each of the proteins based on the binding measurement and the binding probability.

2. The method according to claim 1, wherein the step of characterizing the modified form of the protein includes quantifying the modified form of the protein based on the PTM.

3. The method according to claim 1, wherein the PTM is phosphorylated.

4. The method according to claim 1, wherein the PTM is phosphorylated or ubiquitinated.

5. The method according to claim 1, wherein each of the affinity reagents is an antibody or a fragment thereof.

6. The method according to claim 1, wherein the step of characterizing the modified form of the protein includes identifying an isoform of the protein.

7. The method according to claim 1, wherein the step of characterizing the modified form of the protein includes quantifying the isoform of the protein.

8. It is a system, A system including one or more processors and memory, configured to do the following: (a) Receiving a binding measurement indicating the binding outcome of a set of affinity reagent probes applied to a protein in a sample, wherein each of the proteins is immobilized at a specific spatial position on a substrate, and the set of affinity reagents comprises a plurality of individual affinity reagents, each of which has known binding properties to different epitopes, including post-translational modifications (PTMs); (b) receiving the binding probability of each of the individual affinity reagent probes to different epitopes including the PTM; and (c) Characterizing the modification morphology of the protein in the sample by identifying the PTM in the epitope for each of the proteins based on the binding measurement and the binding probability.

9. The system according to claim 8, wherein characterizing the modified form of the protein includes quantifying the modified form of the protein based on the PTM.

10. The system according to claim 8, wherein the PTM is phosphorylated.

11. The system according to claim 8, wherein the PTM is phosphorylated or ubiquitinated.

12. The system according to claim 8, wherein each of the affinity reagents is an antibody or a fragment thereof.

13. The system according to claim 8, wherein characterizing the protein includes identifying isoforms of the protein.

14. The system according to claim 8, wherein characterizing the protein includes quantifying the isoforms of the protein.

15. A computer program product comprising a non-transient, computer-readable medium on which computer program instructions are stored, wherein the computer program instructions are configured to cause one or more computing devices to perform the following actions when executed by one or more computing devices: (a) A binding measurement is received that shows the binding outcome of a set of affinity reagent probes applied to the protein of the sample, where each of the proteins is immobilized at a specific spatial position on the substrate, and the set of affinity reagents comprises a plurality of individual affinity reagents, each of which has known binding properties to different epitopes, including post-translational modifications (PTMs); (b) receive the binding probability of each of the individual affinity reagent probes to different epitopes including the PTM; and (c) The modification morphology of the protein in the sample is characterized by identifying the PTM in the epitope for each of the proteins based on the binding measurement and the binding probability.

16. The computer program product according to claim 15, wherein characterizing the modified form of the protein includes quantifying the modified form of the protein based on the PTM.

17. The computer program product according to claim 15, wherein the PTM is phosphorylated.

18. The computer program product according to claim 15, wherein the PTM is phosphorylated or ubiquitinated.

19. The computer program product according to claim 15, wherein each of the affinity reagents is an antibody or a fragment thereof.

20. The computer program product according to claim 15, wherein characterizing the protein includes identifying isoforms of the protein.

21. The computer program product according to claim 15, wherein characterizing the protein includes quantifying the isoforms of the protein.

Images