Systems and methods for identifying peptides

Abstract

Systems and methods for identifying peptides in a query profile. The methods may include receiving one or more parameters of a query profile; generating one or more candidate peptide sequences based on one or more parameters of the query profile; generating a plurality of samples of the query spectrum; selecting at least one sample from the plurality of samples to compare to the one or more candidate peptide sequences; determining a likelihood indicator for each of the one or more candidate peptide sequences based on a comparison with the at least one sample; applying a signal-to-noise ratio filter to the one or more candidate peptide sequences based on the likelihood indicators of the candidate peptide sequences; and selecting at least one candidate peptide sequence as a proposed peptide sequence in the query spectrum based on the filtered candidate peptide sequences.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS:

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 357,534, filed on June 30, 2022, and U.S. Provisional Patent Application No. 63 / 402,157, filed on August 30, 2022. The entire contents of U.S. Provisional Patent Application No. 63 / 357,534 and U.S. Provisional Patent Application No. 63 / 402,157 are incorporated herein by reference. Technical field:

[0002] The embodiments are directed to mass spectrometry systems and methods. In some example embodiments, the systems and methods may be directed to the identification of peptides. Background technology:

[0003] Biological drugs such as antibodies or nanobodies are currently the fastest growing area in drug development. However, while antibodies are large proteins, only the hypervariable binding sites of antibodies can recognize and bind to drug targets, receptors or ligands. Creating monoclonal antibodies, single-chain antibodies or nanobodies with a single binding site is a complex problem that requires cloning and expressing antibody variable domains that are synthesized and secreted from cells or expressed on the surface of viral phages, which is both expensive and time-consuming.

[0004] Peptide drugs have enormous health and economic value. Globally, many peptide drugs are in use, in clinical development or in preclinical studies. In recent years, several peptide drugs have been approved by the U.S. Food and Drug Administration (FDA), and the proportion of peptide drugs in all FDA-approved drugs is expected to continue to increase due to breakthroughs in peptide synthesis, peptide drug stabilization and delivery. Developing an efficient method to identify peptide drugs, namely peptide aptamers, will bring great benefits to mankind, however, it is a particular challenge due to the difficulty in reliably isolating and characterizing peptide aptamers that bind to drug targets. Summary of the invention:

[0005] Various embodiments described herein are generally directed to computer-implemented methods (and related systems configured to implement the methods) for identifying peptides in a query spectrum.

[0006] An exemplary computer-implemented method may include receiving one or more parameters of a query spectrum; generating one or more candidate peptide sequences based on the one or more parameters of the query spectrum; generating multiple samples of the query spectrum; selecting at least one sample from the multiple samples for comparison with the one or more candidate peptide sequences; determining a likelihood index for each of the one or more candidate peptide sequences based on the comparison with the at least one sample; applying a signal-to-noise filter to the one or more candidate peptide sequences based on the likelihood index of the candidate peptide sequences; and selecting the at least one candidate peptide sequence as a proposed peptide sequence in the query spectrum based on the filtered candidate peptide sequences.

[0007] In some embodiments, at least one sample is selected from multiple samples to be compared with one or more candidate peptide sequences. For each sample, it can include: the number of consecutive b and y spectral lines that matches the sample; and if the number of consecutive b and y spectral lines that matches the sample is less than a predetermined minimum number of spectral lines, excluding the sample from comparison with the one or more candidate peptide sequences.

[0008] In some embodiments, selecting at least one sample from a plurality of samples for comparison with one or more candidate peptide sequences may include, for each sample: determining the spectral intensity of the sample; determining the number of matching b and y spectral lines; for the matching b and y spectral lines, determining the sum of the corresponding spectral intensities; determining the sum of the spectral intensities; determining the ratio of the sum of the spectral intensities of the matching b and y spectral lines to the sum of the spectral intensities; and if the ratio is less than a predetermined threshold, excluding the sample from comparison with the one or more candidate peptide sequences.

[0009] In some embodiments, selecting at least one sample from a plurality of samples to be compared with one or more candidate peptide sequences may include, for each sample: determining a precursor mass of the sample; and if the precursor mass is substantially equal to the mass of the candidate peptide sequence, selecting the sample to be compared with the one or more candidate peptide sequences.

[0010] In some embodiments, the precursor mass can include at least one of a precursor mass having a charge state of 1, 2, or 3.

[0011] In some embodiments, the precursor mass can include a mass shift from one or more post-translational modifications.

[0012] In some embodiments, selecting at least one sample from a plurality of samples to be compared with one or more candidate peptide sequences may include, for each sample: determining the mass of a theoretical ion of the sample; and if the mass of the theoretical ion is substantially equal to the mass of the base peak, selecting the sample to be compared with the one or more candidate peptide sequences.

[0013] In some embodiments, determining a likelihood indicator for each of one or more candidate peptide sequences based on comparison with at least one sample can include using one or more of linear regression, multiple linear regression, nested regression, or a linear model.

[0014] In some embodiments, it can include determining a likelihood indicator for each of one or more candidate peptide sequences based on comparison with the at least one sample, and for each sample, it can include comparing the theoretical ions of the sample with the corresponding theoretical ions of the candidate peptide sequence using a chi square test.

[0015] In some embodiments, determining a likelihood indicator for each of the one or more candidate peptide sequences based on comparison with the at least one sample may include, for each sample, determining a cross correlation score relative to the candidate peptide sequence.

[0016] In some embodiments, applying a signal-to-noise filter to one or more candidate peptide sequences based on the likelihood index of the candidate peptide sequence may include: generating a high-frequency random source noise spectrum; and determining a difference between the high-frequency random source noise spectrum and the one or more candidate peptide sequences. If the difference exceeds a predetermined threshold difference, the method may include excluding the sample from the selection of the proposed peptide sequence in the query spectrum; otherwise, the method may include including the sample as a selection of the proposed peptide sequence.

[0017] In some embodiments, generating the high frequency random source noise spectrum may include using Monte Carlo random simulation.

[0018] In some embodiments, selecting at least one candidate peptide sequence as a proposed peptide sequence in a query spectrum can include performing a best fit for each spectrum at one of a peptide, accession number, or gene symbol level.

[0019] In some embodiments, generating one or more candidate peptide sequences based on one or more parameters of a peptide query may include one or more of the following: storing multiple peptide sequences in a computer-readable medium; selecting at least one of the multiple peptide sequences stored in the computer-readable medium to serve as at least one candidate peptide sequence based on the one or more parameters; and randomly generating at least one candidate peptide sequence based on the one or more parameters.

[0020] In some embodiments, receiving one or more parameters of the query profile may include receiving the one or more parameters from a user input at a computing device. The one or more parameters may include one or more of an amino acid, a position of an amino acid, a group of amino acids, or a position of the group of amino acids.

[0021] In some embodiments, receiving one or more parameters of a query spectrum may include: receiving a query spectrum; and deriving the one or more parameters from the query spectrum. The one or more parameters may include one or more of neutral loss, post-translational modification shift, iminium ion, B ion subtraction, or Y ion subtraction.

[0022] In some embodiments, selecting at least one of the multiple stored peptide sequences from a computer-readable medium based on the one or more parameters to serve as at least one candidate peptide sequence may include: identifying a subset of peptide sequences from the multiple stored peptide sequences, each peptide sequence satisfying the one or more parameters; and selecting the at least one candidate peptide sequence from the subset of peptide sequences.

[0023] In some embodiments, the plurality of stored peptide sequences may include one or more of naturally occurring peptide sequences and synthetic peptide sequences.

[0024] In some embodiments, randomly generating at least one candidate peptide sequence based on the one or more parameters may include: randomly generating a peptide sequence; determining whether the randomly generated peptide sequence satisfies the one or more parameters; and if the randomly generated peptide sequence satisfies the one or more parameters, using the randomly generated peptide sequence as a candidate peptide sequence, otherwise discarding the randomly generated peptide sequence.

[0025] In some embodiments, each of the at least one randomly generated candidate peptide sequence can have a predetermined length.

[0026] In some embodiments, at least one candidate peptide sequence is randomly generated based on the one or more parameters, and for each randomly generated candidate peptide sequence, it can include: assigning at least one amino acid to the randomly generated candidate peptide sequence based on the one or more parameters; randomly selecting an unassigned position of the predetermined length; and for the unassigned position, randomly selecting an amino acid and assigning the amino acid to the unassigned position. The method can also include continuing to randomly assign amino acids to the randomly generated candidate peptide sequence until each position is assigned.

[0027] In some embodiments, randomly selecting an amino acid may include: randomly selecting an amino acid group from a plurality of amino acid groups, each amino acid group comprising a plurality of amino acids; and randomly selecting the amino acid from the randomly selected amino acid group.

[0028] In some embodiments, each randomly generated candidate peptide sequence may include at least one amino acid from each of the plurality of amino acid groups.

[0029] In some embodiments, generating a plurality of samples of a query spectrum may include one or more of generating an experimental spectrum or generating a simulated spectrum.

[0030] In some embodiments, generating a simulated spectrum can include using a Monte Carlo stochastic simulation.

[0031] In another broad aspect, a system for identifying peptides in a query spectrum is disclosed herein. The system may include a non-transitory computer-readable medium and a processor. The processor is operable to: receive one or more parameters of the query spectrum; generate one or more candidate peptide sequences based on the one or more parameters of the query spectrum; generate multiple samples of the query spectrum; select at least one sample from the multiple samples to compare with one or more candidate peptide sequences; determine a likelihood index for each of the one or more candidate peptide sequences based on the comparison with the at least one sample; apply a signal-to-noise filter to the one or more candidate peptide sequences based on the likelihood index of the candidate peptide sequences; and select at least one candidate peptide sequence as a proposed peptide sequence in the query spectrum based on the filtered candidate peptide sequences.

[0032] In some embodiments, the processor can operate for each sample: the number of consecutive b and y spectral lines that match the sample; and if the number of matching consecutive b and y spectral lines of the sample is less than a predetermined minimum number of spectral lines, the sample is excluded from comparison with the one or more candidate peptide sequences.

[0033] In some embodiments, the processor can perform the following operations for each sample: determine the spectral intensity of the sample; determine the number of matching b and y spectral lines; for the matching b and y spectral lines, determine the sum of the corresponding spectral intensities; determine the sum of the spectral intensities; determine the ratio of the sum of the spectral intensities of the matching b and y spectral lines to the sum of the spectral intensities; and if the ratio is less than a predetermined threshold, exclude the sample from comparison with the one or more candidate peptide sequences.

[0034] In some embodiments, the processor may perform the following operations for each sample: determine the precursor mass of the sample; and if the precursor mass is substantially equal to the mass of the candidate peptide sequence, select the sample to be compared with the one or more candidate peptide sequences.

[0035] In some embodiments, the precursor mass can include at least one of a precursor mass having a charge state of 1, 2, or 3.

[0036] In some embodiments, the precursor mass can include a mass shift from one or more post-translational modifications.

[0037] In some embodiments, the processor may perform the following operations for each sample: determine the mass of a theoretical ion of the sample; and select the sample for comparison with the one or more candidate peptide sequences if the mass of the theoretical ion is substantially equal to the mass of the base peak.

[0038] In some embodiments, the processor may perform the following operations for each sample: determine a likelihood index for the sample using one or more of linear regression, multivariate linear regression, or nested regression based on comparison with the one or more candidate peptide sequences.

[0039] In some embodiments, the processor may perform the following operation for each sample: compare the theoretical ions of the sample with the corresponding theoretical ions of the candidate peptide sequence using a chi-square test to determine a likelihood index for the sample.

[0040] In some embodiments, the processor may perform the following operations for each sample: determine a cross-correlation score relative to the candidate peptide sequence to determine a likelihood indicator for each of the at least one sample based on comparison with the one or more candidate peptide sequences.

[0041] In some embodiments, the processor may perform the following operations: generate a high-frequency random source noise spectrum; and determine the difference between the high-frequency random source noise spectrum and the one or more candidate peptide sequences. If the difference exceeds a predetermined threshold difference, the processor may exclude the sample from the proposed peptide sequence selection in the query spectrum; otherwise, the processor may include the sample in the proposed peptide sequence for selection.

[0042] In some embodiments, the processor may perform the following operations: using Monte Carlo stochastic simulation to generate a high frequency random source noise spectrum.

[0043] In some embodiments, the processor can perform the following operations: best fit each spectrum at one of the level of peptide, accession number or gene symbol.

[0044] In some embodiments, the processor can perform the following operations: store multiple peptide sequences in a computer-readable medium; based on the one or more parameters, select at least one of the multiple peptide sequences stored in the computer-readable medium to be used as at least one candidate peptide sequence among the one or more candidate peptide sequences; and based on the one or more parameters, randomly generate at least one candidate peptide sequence among the one or more candidate peptide sequences.

[0045] In some embodiments, the processor may perform the following operations: receiving the one or more parameters input by a user of a computing device, the one or more parameters comprising one or more of an amino acid, a position of an amino acid, an amino acid group, or a position of the amino acid group.

[0046] In some embodiments, the processor may perform the following operations: receiving the query spectrum; and obtaining the one or more parameters from the query spectrum. The one or more parameters may include one or more of neutral loss, post-translational modification shift, iminium ion, B ion subtraction, or Y ion subtraction.

[0047] In some embodiments, the processor may perform the following operations: identifying a subset of peptide sequences from a plurality of stored peptide sequences, each peptide sequence satisfying the one or more parameters; and selecting the at least one candidate peptide sequence from the subset of peptide sequences.

[0048] In some embodiments, the plurality of stored peptide sequences may include one or more of naturally occurring peptide sequences and synthetic peptide sequences.

[0049] In some embodiments, the processor can perform the following operations: randomly generate a peptide sequence; determine whether the randomly generated peptide sequence satisfies the one or more parameters; and if the randomly generated peptide sequence satisfies the one or more parameters, use the randomly generated peptide sequence as a candidate peptide sequence, otherwise discard the randomly generated peptide sequence.

[0050] In some embodiments, each of the at least one randomly generated candidate peptide sequence can have a predetermined length.

[0051] In some embodiments, the processor may perform the following operations: assigning at least one amino acid to the randomly generated candidate peptide sequence based on the one or more parameters; randomly selecting an unassigned position of the predetermined length; and for the unassigned position, randomly selecting an amino acid and assigning the amino acid to the unassigned position. The processor may continue to randomly assign amino acids to the randomly generated candidate peptide sequence until each position is assigned.

[0052] In some embodiments, the processor may perform the following operations: randomly select an amino acid group from a plurality of amino acid groups, each amino acid group comprising a plurality of amino acids; and randomly select the amino acid from the randomly selected amino acid group.

[0053] In some embodiments, each randomly generated candidate peptide sequence can include at least one amino acid from each of a plurality of amino acid groups.

[0054] In some embodiments, the processor may be operable to generate an experimental spectrum or a simulated spectrum.

[0055] In some embodiments, the processor may be operable to generate the simulated spectra using Monte Carlo stochastic simulation. Description of the drawings:

[0056] Several embodiments will now be described in detail with reference to the accompanying drawings, in which: Figure 1 is a block diagram of components of a peptide identification system according to an example embodiment; Figure 2A is a flow chart of a method for identifying peptides in a query spectrum according to an example embodiment; Figure 2B is a schematic diagram of a method for identifying peptides in a query spectrum according to an example embodiment; Figure 3A is a table of amino acid groups according to example embodiments; Figure 3B is a table of possible combinations of various predetermined lengths according to an example embodiment, wherein each amino acid group has at least one amino acid; Figure 4A is a diagrammatic representation of randomly generated candidate peptides according to an example embodiment; Figure 4B is a graphical representation of randomly generated candidate peptides according to another example embodiment; Figure 5A is a flow chart of selecting a sample to compare with a candidate peptide sequence selected from a memory according to an example embodiment; Figure 5B is a flow chart for selecting a sample for comparison with a randomly generated candidate peptide sequence according to an example embodiment; Fig. 6A is a graphical representation of a spectrum according to an example embodiment; Figure 6B is a graphical representation of m / z values ​​of observed and expected spectral lines for matching ions according to an example embodiment; Figure 6C is a graphical representation of possible relationships between peptides, accession numbers and gene symbols according to example embodiments; Fig. 7A is a table of likelihood indicators of various candidate peptide sequences selected from the memory for a given query spectrum according to an example embodiment; Figure 7B is a table of likelihood indicators of various candidate peptide sequences selected from the memory for a given query spectrum according to an example embodiment; Figure 7C is a table of likelihood indices for various candidate peptide sequences in a given query spectrum taking into account charge states, according to an example embodiment; Fig. 8A is a graphical representation of results from 18 standard protein tests of a peptide generation system according to an example embodiment; Figure 8B Schematic diagram of the test results of 18 standard proteins by the XITandem system; Figure 8C Shows Results of 18 standard protein tests using the Proteome Discoverer system; Fig. 9A is a graphical representation of the results of testing 18 standard proteins using the peptide generation system of the bos Taurus protein library according to an exemplary embodiment; Fig. 9B It is a graphic representation of the results of testing 18 standard proteins using the XITandem system of the bos Taurus protein library; Fig. 9C The bos Taurus protein library is shown. The results of the Proteome Discoverer system testing 18 standard proteins; Fig. 10A , 10B and 10C were calculated using three different thresholds in peptide aptamers, XITandem and Overlapping gene symbols identified between; Fig.11A is a graphical representation of a linear regression model fitting a query spectrum to various candidate peptide sequences according to an example embodiment; Fig. 11B yes Fig.11A The data table of the linear regression model; Fig. 12A is a table showing the results of different fitting methods for the EGVNDNEEGFFSAR glufib peptide searched against a small cow library and narrowed to the top 30 gene symbols using filters, according to an example embodiment; and Fig. 12Bis a table showing the results of different fitting methods for the EGVNDNEEGFFSAR glufib peptide, searched against a small cow library and not narrowed down to the top 30 gene symbols using filters, according to an example embodiment.

[0057] The accompanying drawings described below are intended to illustrate aspects and features of various examples of the embodiments described in the present application, but are not intended to limit these aspects and features. For simplicity and clarity of explanation, the elements shown in the drawings are not necessarily drawn to scale. For clarity, the sizes of certain elements may be enlarged relative to other elements. It should be understood that for simplicity and clarity of explanation, reference numerals may be repeated in the various figures to indicate corresponding or similar elements or steps, where appropriate. Specific implementation method:

[0058] Reliable identification of peptide sequences from complex samples by mass spectrometry remains difficult due to the huge number of possible amino acid and post-translational modification combinations.

[0059] There are two common strategies for matching spectra to peptides: database search methods, where protein libraries exist, and de novo sequence generation methods, where libraries are not required. Database search methods (also called library searches) are more common. Using database search methods, MS / MS spectra are matched to known peptides stored in a database. Database searches require a deep understanding of the possible peptides in the sample to create a theoretical peptide list and restrict the discovery space to known peptides.

[0060] Specifically, bait library search involves searching for existing libraries connected to non-existent (i.e., fictitious) peptides with similar length and amino acid distribution. However, typically, the spectrum matched by the bait library search method is less than half of all the spectra collected. Studies have found that the low matching degree of the bait library search method can be largely attributed to the assignment of the MS / MS spectrum of the actual peptide to the peptide that does not exist in the connected bait, and the combination of post-translational modifications is not considered. It has been shown that peptides can be fitted only by MS / MS fragmentation spectra, and peptide mass changes of plus or minus hundreds of Daltons can be allowed to capture all post-translational modifications. Quality tolerance database search identifies a large part of the unassigned spectra in shotgun proteomics as modified peptides (Nat Biotechnol.33:743-9), but the method is computationally expensive. In this regard, the cost and efficiency of computer search are increasing rapidly, and new algorithms may be needed to calculate false positive rates and new signal-to-noise filters.

[0061] Other studies have pointed out that the matching is poor because the combination of post-translational modifications is not taken into account. The poor matching is also a result of the decoy library search strategy itself. It assumes low degrees of freedom because the raw data are spectra and the data are almost continuous.

[0062] Library search methods rely on well-annotated libraries. Therefore, database search methods are limited to samples with adequately annotated peptide libraries. Even in a well-annotated library, a peptide corresponding to a particular MS / MS may not exist. The peptide may be unannotated, may be from a different species, or may result from a genetic mutation. The present invention aims to facilitate the identification of drug target binding peptides in the absence of a peptide library.

[0063] De novo methods aim to identify peptide sequences using a series of rules without the use of peptide libraries and are usually computationally expensive.

[0064] Traditional de novo sequencing methods aim to identify peptide sequences without using peptide libraries of possible amino acid sequences, which results in a huge search space, which is made even larger by the large number of possible post-translational modifications. De novo methods can be biased, so a probability-based mechanism is needed to select the correct sequence from many possible de novo amino acid sequences. Probabilistic mechanisms may involve calculating goodness of fit or correlation or linear models, but may also be based on the observed frequency of random (i.e., noise) MS / MS spectra generated randomly in computer or physically.

[0065] Disclosed herein are systems and methods for identifying peptides from MS / MS spectra, wherein the peptides can be retrieved from a known annotated library or generated using a random peptide generator. The random generation of peptides can be performed at runtime. In addition, a combination of user-imposed constraints and de novo rules of thumb can be applied to narrow down the possible combinations of peptide sequences. Also disclosed are systems and methods for creating synthetic peptide libraries, wherein peptide length and amino acid groups are selectable. These features allow the creation of a list of peptides with desired characteristics, which can reduce the search space of the random peptide generator. The systems and methods can be used to identify peptide drugs and / or aptamers.

[0066] Therefore, a straightforward method is needed to isolate peptide binding regions that interact directly with drug target receptors, ligands, and diagnostic or prognostic proteins.

[0067] In some embodiments, the methods and systems disclosed herein involve screening naturally occurring and / or synthetic peptide libraries by direct binding to immobilized drugs, drug targets, ligands, receptors, markers or other proteins or biopolymers or biomolecules or biochemicals or chemical drugs or portions thereof, hereinafter referred to as "binding targets using LC-ESI-MS / MS."

[0068] The methods and systems disclosed in the present application can use real physical libraries of naturally occurring and / or randomly synthesized peptides and / or polypeptides. Real physical libraries in the present application refer to a plurality of peptides that can be used in experiments, such as dumping on binding targets.

[0069] The methods and systems disclosed herein may include the use of samples prepared according to known methods. For example, the sample may be prepared by: (1) immobilizing the binding target on a surface; (2) allowing the immobilized binding target to interact with a natural or synthetic polypeptide library; (3) washing away unbound polypeptides; and (4) eluting the bound polypeptides, for example, by eluting with salts or surfactants or acids or bases or organic solvents or chaotropes or any combination thereof. The methods and systems also include analyzing the bound, washed and eluted polypeptides by LC-ESI-MS / MS or LC-ESI-MSN directly or after chemical or enzymatic modification. In some embodiments, MS / MS or MSn may be fitted to peptide sequences with or without a database, or the peptides may be fitted to randomly generated or other peptides using statistical methods. Optionally, the methods and systems may also include using the principles of peptide library design and de novo sequencing to constrain and guide the creation of a peptide library based on amino acid composition determined from MS, MS / MS and / or MSn. The methods and systems disclosed in the present application may also include: generating a computer peptide library to match the characteristics of the physical library used in the experiment, and using the amino acids observed from the MS / MS or MSn spectrum to reduce the dimension of each MS / MS or MSn spectrum searched; using ab initio, goodness of fit, regression or linear model or correlation or heuristic algorithm to fit the observed MS / MS or MSn spectrum of the physical peptide binding target with the possible peptides in the computer library to determine the amino acid sequence of the physical peptide binding target and the molecular composition of the substance; synthesizing and testing specific recognized peptides to bind to the target molecule for use as a drug, ligand, receptor or diagnostic or prognostic affinity binding agent. The methods and systems disclosed in the present application can be iterative, where the peptides obtained from the first round of the process are used to guide the modification of the physical peptide library for subsequent refinement. The systems and methods disclosed in the present application can be used to create new biopharmaceuticals (i.e., peptide aptamers) without the cloning and expression steps currently used to create monoclonal antibodies, nanobodies or phage binding sites.

[0070] The methods and systems disclosed in the present application may include: cloning and covalently immobilizing targets onto high-throughput 96-well PVDF plates or micro / nano surfaces or labeled targets in solution; applying known or unknown natural or synthetic peptide libraries for binding, washing, elution (with salts, organics, acids, bases, competitors, analogs or labeled compounds), and finely separating samples into discrete fractions by nanospray liquid chromatography (e.g., MALDI, electrospray or other ionization methods), and performing MSn fragmentation by gas collision or impact, chemical or electronic impact or electrical energy or acceleration or atomic collision or interaction with other particles; and using search engines and proteomics platforms that utilize key aspects of library searching and random peptide generators to identify peptide drugs.

[0071] The method and system disclosed in the present application can match the spectrum of a sample with a peptide sequence. In some embodiments, the search engine can use a combination of scoring algorithms, such as a fitting algorithm based on the number of spectral lines, the number of sequential ions, and chi-square, linear regression, amino acid matching, and signal-to-noise ratio intensity scores to identify peptide sequences. The search engine can switch back and forth between the spectrum generated by the reference library and the random number generator given certain parameters according to the nature of the experiment. The search engine uses a series of nested loops, and fragments can be mapped to their respective precursor masses to perform multi-stage fragmentation events (MSn). The disclosed embodiment can verify the results by, for example, calculating the number of spectra that match each identified peptide sequence in the sample and comparing the number of (e.g., blank noise) spectra obtained by random simulation and / or experiment.

[0072] Now refer to Figure 1 , which shows a block diagram 100 of components that interact with a peptide recognition system 110. Figure 1 As shown, peptide aptamer recognition system 110 communicates with computing device 120 and external data storage 130 via network 140 .

[0073] The peptide identification system 110 includes a processor 112, a communication component 114, and a (related) data storage component 116. The peptide identification system 110 may be provided on one or more computer servers, which may be distributed over a wide geographic area and connected via a network 140.

[0074] The processor 112, the communication component 114, and the data storage component 116 can be combined into a smaller number of components or separated into more components. In some cases, one or more components may be distributed over a wide geographic area. The processor 112, the communication component 114, and the data storage component 116 can be implemented in software or hardware or a combination of software and hardware.

[0075] The processor 112 can implement the operation of controlling the peptide identification system 110. The processor 112 can start and manage the operation of each other component in the peptide identification system 110. The processor 112 can be any suitable processor, controller or digital signal processor, graphics processing unit, application specific integrated circuit (ASIC) and / or field programmable gate array (FPGA), which can provide sufficient processing power according to the configuration, purpose and requirements of the peptide identification system 110. In some embodiments, the processor 112 may include multiple processors, each processor is configured to perform different dedicated tasks. The processor 112 can execute various instructions stored in the data storage 116 to implement various methods described in the present application.

[0076] The communication component 114 may include any interface that enables the peptide identification system 110 to communicate with other devices and systems. In some embodiments, the communication component 114 may include at least one of a serial port, a parallel port, or a USB port. The communication component 114 may also include a One or more of WiFi, Internet, local area network (LAN), Ethernet, Firewire, modem, fiber optic, or digital subscriber line connections interface with the components. Various combinations of these elements can be incorporated into the communication component 114. For example, the communication component 114 can be used to communicate with the computing device 120 to receive the query MS / MS spectrum and one or more parameters of the query MS / MS spectrum.

[0077] For example, the communication component 114 can receive input from various input devices, such as a mouse, keyboard, touch screen, thumb wheel, trackpad, trackball, card reader, voice recognition software, etc., depending on the requirements and implementation of the peptide identification system 110.

[0078] The data storage component 116 may include RAM, ROM, one or more hard disk drives, one or more flash drives, or some other suitable data storage element (e.g., a disk drive, etc.). The data storage component 116 may include one or more databases for storing information related to peptide sequences, amino acids, and amino acid groups. The data storage 116 may also store data for performing one or more tests, such as, but not limited to, precursor tests and base peak tests. The data storage 116 may also store instructions executable by the processor 112 to implement the various methods described herein.

[0079] Similar to the data storage component 116, the external data storage 130 may also include a RAM, a ROM, one or more hard disk drives, one or more flash drives, or some other suitable data storage element (e.g., a disk drive, etc.). In some embodiments, the external data storage 130 may be similar to the data storage 116, but is located away from the peptide identification system 110 and can be accessed via the network 140. The external data storage 130 may include one or more databases for storing information related to peptide sequences, amino acids, and amino acid groups. In addition, the external data storage 130 may also store data for performing one or more tests, such as, but not limited to, precursor tests and base peak tests. Similar to the data storage 116, the external data storage 130 may store various data related to the operation of the peptide identification system 110.

[0080] Computing device 120 may include any networked device that can connect to network 140. A networked device is a device that is capable of communicating with other devices over a network, such as network 140. A networked device may be coupled to network 140 via a wired or wireless connection. Figure 1 Only one computing device 120 is shown in FIG. 1 , but it is understood that more computing devices 120 may be connected to the network 140 .

[0081] The computing device 120 may include at least a processor and a memory, and may be an electronic tablet device, a personal computer, a workstation, a server, a portable computer, a mobile device, a personal digital assistant, a laptop computer, a smart phone, a WAP phone, an interactive television, a video display terminal, a game console, a portable electronic device, or any combination of these.

[0082] Network 140 may be any network capable of carrying data, including the Internet, Ethernet, plain old telephone service (POTS) lines, public switched telephone network (PSTN), integrated services digital network (ISDN), digital subscriber line (DSL), coaxial cable, fiber optic, satellite, mobile, wireless (e.g., Wi-Fi, WiMAX, ultra-wideband, ), SS7 signaling network, fixed lines, local area networks, wide area networks and other networks, including any combination of these networks, can interface with the peptide recognition system 110, the computing device 120 and the external data storage 130 and enable communication between them.

[0083] It should be understood that Figure 1 Some components in the system, such as components of the peptide identification system 110 or the external data storage 130, can be implemented in a cloud computing environment.

[0084] Now refer to Figure 2A , which shows a flow chart of a method 200A for identifying peptides in a query spectrum according to some embodiments. Figure 2B , which shows a schematic diagram of a method 200B for identifying peptides in a query spectrum according to some embodiments. To help describe Figure 2A-2B The methods 200A and 200B shown in FIG. Figures 3A to 6B A peptide identification system (eg, peptide identification system 110 having processor 112) may be configured to implement method 200. A query profile may correspond to a sample.

[0085] Method 200 may begin at step 210, when a processor (eg, processor 112) may receive one or more parameters of a query profile. Processor 112 may receive the one or more parameters from a computing device (eg, computing device 130) via a network (eg, network 140).

[0086] In some embodiments, one or more parameters may be received from a user input at the computing device 130. For example, the one or more parameters may be one or more of an amino acid or an amino acid group. In addition, the position corresponding to each amino acid or amino acid group may be included.

[0087] In some embodiments, the processor 112 can receive a query spectrum and one or more parameters obtained from the query spectrum. For example, the one or more parameters can be one or more of neutral loss (including ammonia loss and water loss), post-translational modification shift (may only be located on a small group of amino acids), iminium ion (e.g., methionine oxidation), subtraction of B ions, or subtraction of Y ions.

[0088] In step 212, the processor 112 can generate one or more candidate peptide sequences based on one or more parameters. In some embodiments, generating one or more candidate peptide sequences can include selecting at least one of a plurality of peptide sequences stored in the memory from the memory to be used as at least one candidate peptide sequence. A plurality of peptide sequences can be stored in a memory, such as a data storage component 130 or a data storage component 116. For example, a plurality of peptide sequences can be stored in a peptide database or library. In some embodiments, a plurality of peptide sequences can include naturally occurring peptide sequences and / or synthetic peptide sequences. When the natural aptamer is from a known source and / or the possible peptide sequence is known (for example, but not limited to digested plasma), it may be appropriate to select a candidate peptide sequence from the memory.

[0089] The selection of the candidate peptide sequence from the memory may be based on one or more parameters received at step 210, which may be used as constraints to narrow the search space of the candidate peptide sequences. For example, the one or more parameters may include specific amino acids (e.g., lysine or arginine) of the candidate peptide sequence. The processor 112 may identify a peptide sequence having the specific amino acid from a plurality of peptide sequences stored in the memory, and then select the candidate peptide sequence from the peptide sequences having the specific amino acid.

[0090] In another example, one or more parameters may also include a specific position (e.g., last position) of a specific amino acid. The processor 112 may identify a peptide sequence having a specific amino acid at a specific position from a plurality of peptide sequences stored in the memory, and then select a candidate peptide sequence from the peptide sequences having the specific amino acid at the specific position.

[0091] In another example, the one or more parameters may include at least one amino acid group (e.g., basic, acidic, neutral, aliphatic, or polar amino acids) of the candidate peptide sequence. Each amino acid group may include a plurality of amino acids. Similarly, the processor 112 may identify a peptide sequence having amino acids in an amino acid group from a plurality of peptide sequences stored in the memory, and then select a candidate peptide sequence from the peptide sequences having amino acids in the amino acid group.

[0092] In some embodiments, the one or more parameters may include at least one specific amino acid and at least one specific amino acid group. Other constraints are also possible.

[0093] In yet another example, the one or more parameters may also include a specific position (e.g., the last position) of an amino acid of a specific amino acid group. The processor 112 may identify a peptide sequence having an amino acid of a specific amino acid group at a specific position from a plurality of peptide sequences stored in the memory, and then select a candidate peptide sequence from the peptide sequences having an amino acid of a specific amino acid group at a specific position.

[0094] In some embodiments, generating a plurality of candidate peptide sequences may include randomly generating at least one candidate peptide sequence. Randomly generating candidate peptide sequences may be applicable to synthetic peptide sequences or unknown peptide aptamers.

[0095] In at least one embodiment, the processor 112 can randomly generate a peptide sequence and determine whether the randomly generated peptide sequence satisfies one or more parameters. If the randomly generated peptide sequence satisfies one or more parameters, the processor 112 can use the randomly generated peptide sequence as a candidate peptide sequence. However, the search space may be large. In contrast, the search space of the randomly generated sequence may be limited by processing power. In some embodiments, supercomputers or quantum computing can be used to implement this processing in a timely manner.

[0096] In some embodiments, the one or more parameters received in step 210 can be used as constraints to narrow down the possible combinations of randomly generated candidate peptide sequences. For example, the one or more parameters can include a predetermined length of the randomly generated candidate peptide sequences. In the case of a predetermined length of 5, the possible combinations of randomly generated candidate peptide sequences are 3.2 million. Generally, identifying longer peptides requires more processing power.

[0097] Additional constraints can be used to further narrow down the possible combinations of randomly generated candidate peptide sequences. One or more parameters can relate to the minimum number of amino acids in a specific amino acid group, a specific position of amino acids in a specific amino acid group, a specific amino acid and a specific position of a specific amino acid, neutral loss, post-translational modification shift, iminium ions, subtraction of B ions, or subtraction of Y ions.

[0098] In some embodiments, the processor 112 may assign at least one amino acid to the randomly generated candidate peptide sequence according to one or more parameters. For example, the processor 112 may assign a specific amino acid to a specific position according to one or more parameters. Then, the processor 112 may randomly assign amino acids to the remaining unassigned positions.

[0099] Now refer to Figure 3A, which shows a table 300 of different amino acid groups 302. The amino acid groups can include, for example, the anchor point of a tryptic peptide, acidic, basic, polar uncharged, polar uncharged with an amine side chain, sulfide R, neutral, aliphatic, and proline. Each amino acid group can include any number of amino acids, as shown in 304. For example, the anchor point of a tryptic peptide can include two amino acids: arginine and lysine, represented by 306. Acidic amino acids can include two amino acids: aspartic acid and glutamic acid. Basic amino acids can include arginine, histidine, and lysine. Polar uncharged amino acids can include serine and threonine. Polar uncharged amino acids with an amine side chain can include asparagine and glutamine. Sulfide R amino acids can include cysteine ​​and methionine. Neutral amino acids can include alanine, valine, isoleucine, and leucine. Aliphatic amino acids can include phenylalanine, tyrosine, and tryptophan. Finally, the proline amino acid can include proline itself. In some embodiments, the proline amino acid group can be a special case.

[0100] In some embodiments, each randomly generated candidate peptide sequence can include at least one amino acid from each amino acid group. That is, the minimum number of amino acids in each amino acid group is one. Requiring each randomly generated candidate peptide sequence to include at least one amino acid from each amino acid group can significantly reduce possible combinations. Figure 3B Table 320 of possible combinations of various predetermined lengths with at least one amino acid from each amino acid group is shown. In some embodiments, peptides of length 11 or 12 may be used, where at least one amino acid from each amino acid group is required.

[0101] In some embodiments, randomly generating a candidate peptide sequence of a predetermined length may include randomly selecting an unassigned position within the candidate peptide sequence. For the unassigned position, the processor 112 may randomly select an amino acid group from a plurality of amino acid groups and randomly select an amino acid from the selected amino acid group. The selected amino acid may be assigned to the unassigned position, and the processor 112 may continue until each position is assigned. That is, the processor 112 may randomly select another unassigned position, randomly select an amino acid group from a plurality of amino acid groups, and randomly select an amino acid from the selected amino acid group to assign to the unassigned position.

[0102] When the randomly generated candidate peptide sequence requires at least one amino acid from each amino acid group, the subsequent random amino acid group selection can be performed from the unselected amino acid group. However, other methods can also ensure that the randomly generated candidate peptide sequence contains at least one amino acid from each amino acid group.

[0103] Now refer to Figure 4A, which shows an exemplary random generation 400 of candidate peptide sequences. As shown at 402, a predetermined length is used. Figure 4A As shown, a predetermined length of 12 is used. At 404, the amino acid group constraints are applied. That is, each group is assigned to one position except for the special case group. The special case group has only one amino acid, proline. Proline does not belong to any other category.

[0104] At 406, for each position with an amino acid group constraint, an amino acid from that amino acid group is randomly selected. For each position without an amino acid group constraint, an amino acid from any amino acid group is randomly selected. The randomly selected amino acid is compared to a filter condition. If the randomly selected amino acid does not satisfy the filter condition, another amino acid from the group is selected for the position. If the randomly selected amino acid satisfies the filter condition, the amino acid assigned to the position is retained. Peptides that satisfy the filter condition continue to be generated until no new peptides are generated or until a user-defined amount of time has passed, thereby triggering the position of the constraint to be adjusted and a new peptide to be generated.

[0105] At 408, the group constraints applied at 404 are adjusted. In some embodiments, the adjustment may include selecting a different amino acid from the group for the position. In some embodiments, the group constraints may not be adjusted. At 410, an amino acid is randomly selected for the unassigned position. The randomly generated peptides based on the random positions are placed in the "peptide template box". The deadman timer switch is in operation and is reset each time a new random peptide is generated. If the time period set in the "infinite loop control" group expires, the peptide template is readjusted, similar to re-adjusting a deck of cards. The random generator restarts using this new peptide template. This process is repeated indefinitely, hence the name "infinite loop control".

[0106] Now refer to Figure 4B , which shows another example random generation of candidate peptide sequences 450. As shown at 452, a predetermined length is used. Figure 4B As shown, a predetermined length of 12 is used. At 454, an unallocated position is randomly selected. Figure 4B As shown, all 12 positions are unassigned, and position 3 is randomly selected. At 456, an amino acid group from 420 is randomly selected. Figure 4B As shown, an acidic amino acid group is randomly selected. At 458, an amino acid from the acidic amino acid group is randomly selected. Figure 4B As shown, aspartic acid was randomly selected.

[0107] At 460 , additional unassigned positions are assigned until the candidate peptide sequence includes at least one amino acid from each amino acid group 420 .

[0108] At 462 , additional unassigned positions are assigned until every position of the candidate peptide sequence has been assigned. The randomly selected amino acid assigned at 462 may be from any one of the amino acid set 420 .

[0109] In some embodiments, candidate peptide sequences can be randomly generated to include known amino acids at known or unknown positions. In addition, randomly generated candidate peptide sequences can be customized based on known information about a specific binding target.

[0110] In some embodiments, the processor 112 may determine whether to select a candidate polypeptide sequence from a memory or to randomly generate a candidate polypeptide sequence based on one or more parameters received from the computing device 120 .

[0111] In some embodiments, the one or more parameters may include user-specified parameters for selecting whether to select a candidate peptide sequence from a memory or to generate a candidate peptide sequence by random. In some embodiments, the processor 112 may compare the one or more parameters with certain thresholds and determine whether to select a candidate peptide sequence from a memory or to generate a candidate peptide sequence by random.

[0112] In some embodiments, candidate peptide sequences can be selected from a memory and randomly generated. For example, a set of amino acids can be identified from a protein database, and then candidate peptide sequences can be randomly generated to identify post-translational modifications (PTMs) by including amino acid masses with mass shifts (depending on the nature of the PTMs).

[0113] In some embodiments, the peptide identification system 110 may first select a candidate peptide sequence from a memory. Based on the likelihood index, the peptide identification system 110 may be operated again for the same query spectrum, but the candidate peptide sequence is randomly generated.

[0114] Returning now to FIG. 2 , at step 214 , the processor 112 may generate a plurality of samples of the query spectrum in which the peptide is being identified. In some embodiments, the samples of the query spectrum may be generated experimentally or by simulation. In some embodiments, the simulation may be a Monte Carlo random simulation. Generating a plurality of samples may include generating spectra of the samples, as shown in 214B.

[0115] At step 216, the processor 112 may select at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences generated at step 212. Selecting a sample from the plurality of samples for comparison may reduce the computational burden of fitting a large data set to a large search space. In order to select a sample for comparison, one or more signal processing filters may be applied. The signal processing filters may include minimum spectral number, maximum spectral line filtering, precursor mass testing, base peak testing, and / or delta mass filtering.

[0116] The minimum number of spectra includes discarding spectra with line counts below a given threshold, because there may not be enough information below the line threshold to make a reliable peptide match. For the minimum number of spectra, in some embodiments, the processor 112 can calculate the number of spectral lines of the sample. In some embodiments, the processor 112 can calculate the number of matched continuous ion (e.g., b and y ion) spectral lines. If the number of spectral lines of the sample or the number of matched continuous ion spectral lines is less than a predetermined minimum number of spectral lines, the processor 112 can exclude the sample from comparison with one or more candidate peptide sequences. In some embodiments, a correction is applied to the number of matched continuous ion spectral lines. For example, the ratio of the number of matched continuous ion spectral lines to the total number of spectral lines can be determined, and the predetermined minimum number of spectral lines can correspond to a predetermined minimum ratio. In some embodiments, the predetermined minimum number of spectral lines can be specified by a user. For example, the predetermined minimum number of spectral lines can be a parameter received from the computing device 120.

[0117] Maximum spectral line filtering includes limiting the processor 112 to use the strongest spectrum. For maximum spectral line filtering, the processor 112 can determine the spectral intensity of the sample. In some embodiments, the processor 112 can determine the sum of the spectral intensities of the matching ion spectral lines, determine the sum of the spectral intensities (i.e., the sum of the spectral intensities of all spectral lines), and determine the ratio of the sum of the spectral intensities of the matching ion spectral lines to the sum of the spectral intensities. If the ratio is less than a predetermined threshold, the processor 112 can exclude the sample from comparison with one or more candidate peptide sequences. In some embodiments, the predetermined spectral intensity can be based on a candidate peptide sequence, such as a percentage of the spectral intensity of the candidate peptide sequence. The percentage of spectral intensity can be specified by a user. For example, the percentage of spectral intensity can be a parameter received from the computing device 120. In this way, maximum spectral line filtering can be dynamically applied when the predetermined spectral intensity is based on a candidate peptide sequence. The maximum spectrum is related to the intensity value. Assuming the maximum spectrum value is 50, the engine will only check the 50 strongest spectral lines. This significantly reduces the calculation time because a single MS2 spectrum can contain thousands of spectral lines.

[0118] The precursor mass test involves comparing the precursor mass of the sample to the masses of the candidate peptide sequences. Samples whose precursor masses match the masses of any candidate peptides can be processed further. Samples whose precursor masses do not match any candidate peptides can be disregarded and the next sample evaluated.

[0119] The precursor mass test may include the processor 112 determining the precursor mass of the sample. If the precursor mass is substantially equal to the mass of a candidate peptide sequence in one or more candidate peptide sequences, the processor 112 may select the sample to be compared with the one or more candidate peptide sequences. The processor 112 may consider the precursor mass to be substantially equal to the mass of the candidate peptide sequence within a predetermined error range of the mass of the candidate peptide sequence. In some embodiments, the predetermined error range may be specified by a user. For example, the predetermined error range may be a parameter received from the computing device 120.

[0120] In some embodiments, the processor 112 can determine the precursor masses of the sample at different charge states. For example, the processor 112 can determine the precursor masses of charge states 1, 2, and 3.

[0121] In some embodiments, processor 112 can determine the precursor mass of a sample while taking into account the presence of one or more post-translational modifications (PTMs). For example, the precursor mass can include a mass shift caused by one or more post-translational modifications (PTMs).

[0122] The base peak test involves comparing the theoretical ion mass of the sample to the base peak mass. Samples with theoretical ion masses that match the base peak mass can be processed further. Samples with theoretical ion masses that do not match the base peak mass can be removed from further consideration and the next sample evaluated. Fig. 6A , the highest spectral line is considered the "base peak" and is centered at 200 Daltons along the horizontal axis. In contrast to the subject matter disclosed in the present application, many conventional methods start the identification process from the base peak and look up and down the spectrum to establish identification. In some embodiments disclosed in the present application, the base peak value can be used as part of a rapid pass-fail test. Both the precursor mass and the base peak value are provided directly by the instrument to avoid the computational load of processing the entire set of spectral lines. This allows for rapid testing of random peptides on affordable computing systems.

[0123] The base peak test may include the processor 112 determining a theoretical ion mass for the sample. If the theoretical ion mass is substantially equal to the base peak mass, the processor 112 may select the sample for comparison with one or more candidate peptide sequences.

[0124] In some embodiments, if the precursor mass is not present in the observed spectrum, the precursor mass can be added to the theoretical spectrum. By adding the precursor mass to the theoretical spectrum, the likelihood of identifying the first amino acid in the peptide sequence increases and the amino acid match rate (AAMR) score can be improved. Fragmentation is a two-step process. First, an MS1 ​​scan is performed to determine the precursor mass. Then an MS2 scan is performed. Ideally, this precursor mass will also appear in the MS2 spectrum on the far right of the spectrum. However, in practice, this precursor mass may not be present in the MS2 spectrum because amino acids may be "lost" between the MS1 and MS2 scans. If the precursor mass is not present in the MS2 scan, the precursor mass can be added. The missing amino acid will appear at the beginning of the identified peptide. For example, in testing the angiotensin peptide DRVYIHPF, the first amino acid D is usually missing.

[0125] In some embodiments, a theoretical ion can be added to the theoretical spectrum, i.e., the low end of the spectrum can be added. The theoretical ion can represent the terminal mass. Adding the theoretical ion can improve the peptide match mass ratio score. Similar to injecting the precursor mass into the high end of the spectrum, the terminal mass can be injected into the low end of the spectrum. The terminal mass is not an amino acid, but a chemical structure present at the end of a natural peptide. As discussed elsewhere, there are two fragmentation modes, Y and B, with the Y chain starting at the high end or rightmost side of the spectrum, where the precursor mass can be injected, and the B chain analysis starting at the low end or leftmost side of the spectrum. The terminal mass is not usually captured by the mass spectrometer because its mass is lower than what the instrument can detect. By artificially injecting this terminal mass, this mass can be used as a starting point and looking up in the spectrum for amino acids from the B chain. Using the example of angiotensin, whose peptide sequence is DRVYIHPF, this will identify the last amino acid F, which is the last amino acid to fragment before reaching the real or artificially injected terminal mass ion.

[0126] In some embodiments, the MH values ​​of theoretical peptides under charge states 1, 2, or 3 can be added to the theoretical spectrum. The addition of MH values ​​of theoretical peptides under charge states 1, 2, or 3 can be applied to the peptide matching score. After generating a random peptide, the theoretical spectrum and MH value can be calculated. The MH mass can be compared with the precursor mass to calculate the charge of the peptide. However, regardless of the precursor mass, the mass spectrometer may report the spectrum of the peptide in each of the three charge states in a single MS2 scan. If this is the case, the true 1+ spectrum representing the unknown peptide can usually provide the most available spectrum. In some embodiments, the 1+ spectrum can be used as the main identification signal, and then the 2+ and 3+ spectra can be checked to obtain additional evidence for good identification.

[0127] In some embodiments, the MH of a specific peptide can be injected into the observed spectrum. The addition of the MH of a specific peptide can be used to calculate the peptide matching score.

[0128] In some embodiments, the difference between the theoretical mass and the observed peptide mass can be calculated using the estimated charge of the theoretical peptide, the known theoretical mass of the same theoretical peptide, and the known observed precursor mass. The incremental mass (delta mass) can be used to identify modifications, which are additional chemical elements on the peptide. This is also called the modified mass. For example, phosphorylation is a common post-translational modification with a known mass shift of 79.99 Da; if the incremental mass is about this value, it can be determined that the peptide is phosphorylated. Such modifications are related to the role that the peptide may play in human health, because the presence of this element is shown in the precursor mass, but is not part of the theoretical mass and may not appear in the MS2 spectrum. Therefore, the incremental mass value is highly suggestive of a certain modification. In order to test the modification, the modified mass is subtracted from the precursor mass. If the result of this calculation matches the theoretical mass of a random peptide, and the real MS2 spectrum matches the theoretical spectrum, it can be assumed that the theoretical peptide and the modification exist.

[0129] In some embodiments, the application of the signal processing filter can vary. For example, in some embodiments, only the minimum number of spectra can be used. In other embodiments, the minimum number of spectra, maximum spectral line filtering, continuous ion (e.g., b and y ions) quantity and precursor testing can be used. Other combinations are also possible. In addition, the signal processing filter can also depend on whether the candidate peptide sequence is selected from a memory or randomly generated.

[0130] Now refer to Figure 5A , which shows a flowchart 500 for selecting samples to be compared with candidate peptide sequences selected from memory. As shown in flowchart 500, a theoretical peptide list (i.e., a list of candidate peptide sequences) is contained in an outer loop (step 514) and each sample (e.g., an MS / MS spectrum) is contained in an inner loop. One theoretical peptide or candidate peptide sequence is compared to all available scans at a time (step 502) by a testing and scoring algorithm until the peptide list is exhausted. In this embodiment, each theoretical peptide and spectrum pair must pass a precursor quality test (step 506) and a base peak test (step 508) before being scored by at least one of chi-square, multivariate / linear or nested regression, amino acid match ratio, and ion intensity match ratio. User-defined combinations of these scores can be used to identify the best match for each scan (step 512).

[0131] In other embodiments, the loops may be nested in the opposite manner, that is, one sample is compared to each candidate peptide sequence before another sample is compared to each candidate peptide sequence.

[0132] Now refer to Figure 5B, which shows a flow chart 550 for selecting samples to compare with randomly generated candidate peptide sequences. As shown in flow chart 550, a "first in, first out" (FIFO) queue of candidate peptide sequences can be used to replace the outer loop of flow chart 500. The queue grows as candidate peptide sequences are generated, and the queue shrinks each time a candidate peptide sequence is processed.

[0133] Constraints can be used to narrow down the possible combinations of randomly generated candidate peptide sequences. Constraints can be specified by the user, i.e., parameters received from the computing device 120. Constraints can also be determined from a library search method or amino acid matching. As described above, constraints can involve a predetermined length, a minimum number of amino acids from different amino acid groups, specific amino acids at specific positions, and / or precursor mass.

[0134] In the example shown in flowchart 550, constraints related to precursor mass are used. At step 552, processor 112 can retrieve all samples (e.g., MSn scans) with precursor masses within the tolerance range and create a base peak list from these scans. Processor 112 can randomly generate candidate peptide sequences based on a predetermined length and wildcard (step 554), and then generate MH values ​​for the candidate peptide sequences at each of the three charge states (step 558). Processor 112 can run a precursor test on the candidate peptide sequence (step 562), and if it passes, processor 112 can generate a theoretical spectrum (step 564) and run a base peak test (step 566). If the base peak test is passed, the more computationally intensive AAMR and IIMR tests are run (step 568). If the last two scores are above the minimum threshold, the candidate peptide sequence is considered a match and is stored (step 570).

[0135] For technical reasons, the precursor and base peak tests can be repeated again later in the workflow. Specifically, the precursor hints can be a list of m / z values, so a candidate peptide sequence may not match the parent at the precursor level for a particular scan, and similarly, it may not pass the base peak test.

[0136] Returning now to FIG. 2 , at step 218 , based on the one or more test results, the processor 112 may determine a likelihood indicator for each candidate peptide sequence based on a comparison with at least one sample.

[0137] In some embodiments, determining the likelihood index of each candidate peptide sequence may include applying a scoring technique, a filtering technique, or a combination of the two. For example, a scoring technique may be applied first. The scoring technique may be similar to the signal processing filter applied at step 216, including spectral line filtering.

[0138] The likelihood indicator may alternatively or additionally be determined based on additional filtering techniques, including one or more of a chi-square score, a regression score, and a cross-correlation score. Other scoring functions may also be used to generate the likelihood indicator.

[0139] In some embodiments, the candidate peptide sequence matches can be ranked using a chi-square score. For example, a candidate peptide sequence can be selected as a proposed peptide sequence based on the difference between the observed number of ions and the predicted theoretical number of ions for the candidate peptide sequence.

[0140] The chi-square test can be applied to the m / z values ​​of the observed and expected spectra of all matching ions, as described in Equation 1, where the one with the smallest score is ranked highest (i.e., wins):

[0141] Figure 6B The m / z values ​​of the observed spectra (dashed lines) and expected spectra (solid lines) for matching ions are shown. All expected spectra form base peaks.

[0142] Alternatively, the chi-square score can be calculated by adding the number of expected theoretical ions (e.g., b and y ions) for the expected peptide that fall within the M / z range of the mass spectrometer; adding the number of observed ions that match the expected M / z value within the mass resolution limit of the mass spectrometer, and applying Equation 2.

[0143] In some embodiments, the expected theoretical number of ions can be corrected before applying Formula 2. The correction can be applied to compensate for the M / z range and / or mass error of the mass spectrometer. In some embodiments, applying the correction includes multiplying the expected theoretical number of ions by the total number of ions observed and then dividing the product by the number of mass spectrometer bins. The number of bins can be determined according to Formula 3.

[0144] For example, for a mass spectrometer with a range of 50 to 2000 M / z units and a mass resolution of + / - 0.5 M / z units, the number of bins equals 2000-50 / (0.5-(-0.5)=1950.

[0145] In some embodiments, various regression methods can be used to generate likelihood indicators. Linear regression can be used to generate a simple linear regression model for each candidate peptide sequence to determine the best fit model. Multiple linear regression can also be used to incorporate other explanatory variables (such as intensity) into the model.

[0146] In some embodiments, a cross-correlation may be used to generate a likelihood index. The processor 112 may determine a cross-correlation score for the sample relative to the candidate peptide sequence. For example, a scoring function may measure the similarity of the MS / MS spectrum relative to a theoretical spectrum. The cross-correlation may use a sliding dot product. Although cross-correlation may be computationally expensive, various methods may be employed to reduce the computational cost.

[0147] A method for determining a cross-correlation score may include first assigning intensities to theoretical b and y ions. Then the square root of the observed ion intensities for the b and y ions is calculated. Then the average of the cross-correlations for 500 lag times is calculated and this average is subtracted from the cross-correlation between the observed ions and the theoretical ions at a lag time of 0. Other methods for determining cross-correlation scores may be used, such as those described in Eng, JK, McCormack, AL, & Methods described in Yates, JR (1994). "A method for correlating tandem mass spectrometry data of peptides with amino acid sequences in the protein database. Journal of the American Society for Mass Spectrometry," 5(11), 976-989.

[0148] Other scoring functions such as the ion intensity match ratio (IIMR) or the amino acid match ratio (AAMR) may also be used.

[0149] As described above, the ion intensity match ratio (IIMR) generates a signal-to-noise score for each theoretical peptide match. The ion intensity match ratio can be calculated by dividing the sum of the ion intensities that match the theoretical spectrum by the sum of the ion intensities in the observed spectrum. IIMR values ​​range from 0 to 1, with 1 representing a perfect score, and correct for duplicate ion matches.

[0150] It is worth noting that the signal processing filters mentioned above may have reduced the noise. Since the ion intensity match rate is determined after the signal processing filters are applied, the ion intensity match rate can be based on the total ion current (TIC) of the spectrum. That is, the IIMR can be determined by dividing by the TIC of the spectrum.

[0151] By comparing the m / z differences in the observed spectral line pairs, the amino acid match ratios can be used to generate an ordered list of suspected amino acids in the peptide.

[0152] The peptide match rate is calculated by dividing the sum of correctly ordered amino acids by the total number of amino acids used to generate the theoretical spectrum. Because the peptide match rate score is expensive to calculate, the peptide match rate may be calculated only for the best scoring results.

[0153] The independent scoring algorithm can also be used to generate a single vector, using Equation 4. A single vector is more computationally efficient.

[0154] Neutral loss ions and A ions can be considered diagnostic or confirmatory ions that contribute to the score. Ion pairs for identifiable amino acids as described in Peptide Match ions can represent secondary scoring mechanisms associated with these confirmatory ions.

[0155] At step 220, processor 112 may apply a signal-to-noise filter to one or more candidate sequences based on the likelihood indicators of the candidate peptide sequences.

[0156] The mass spectrometer will continuously generate source noise spectra even in the absence of sample. A signal-to-noise filter can be applied using Monte Carlo random simulation to separate the real data from the random spectra by chi-square or other statistical methods. The signal-to-noise filter can eliminate samples that are minimally significant compared to the random or source noise spectra. Random source noise can be generated by applying a blank solvent to the original column or using a random number generator. Peptide sequences that match high-frequency random MS / MS spectra compared to real data by chi-square or other statistical methods are not considered for further analysis.

[0157] In some embodiments, the processor 112 can generate a high frequency random source noise spectrum. The processor 112 can determine the difference between the high frequency random source noise spectrum and one or more candidate peptide sequences. If the difference exceeds a predetermined threshold difference, the sample can be excluded from the proposed peptide sequence in the query spectrum. Otherwise, the sample can be selected as a proposed peptide sequence.

[0158] At step 222, the processor 112 may select a candidate peptide sequence as a proposed peptide sequence in the query spectrum based on the filtered candidate peptide sequences. The proposed peptide sequence may be selected from the filtered candidate peptide sequences based on the best fit. The fit may be associated with a peptide, accession number, or gene symbol. In some embodiments, a machine learning model may be used to select the filtered candidate peptide sequence that best matches the proposed peptide sequence. Figure 2B As shown, in some embodiments, a protein 250 comprising an identified peptide can be identified.

[0159] Although in Figure 2AIn the embodiment, step 220 is shown as being before step 222, but in some embodiments, step 220 can be omitted and the processor 112 can select a candidate peptide sequence from the candidate peptide sequences filtered using a fitting algorithm (e.g., chi-square, linear regression, cross-correlation, etc.) as the proposed peptide sequence in the query spectrum. In such an embodiment, the processor 112 can apply a signal-to-noise ratio filter similar to step 210 to the proposed peptide sequence. For example, the processor 112 can calculate the number of spectra in the sample that match the proposed peptide sequence and compare the number to the number of randomly generated and / or experimentally obtained spectra (e.g., virtual). The processor 112 can exclude samples associated with a number similar to the virtual sample from further analysis.

[0160] The challenge of mass spectrometry analysis is to generate non-redundant data sets for analysis. This challenge can arise from the nature of how peptides, accession numbers, and gene symbols map to each other. Figure 6C An example map 610 showing the relationship between peptides, accession numbers, and gene symbols. Figure 6C As shown, MS / MS spectrum 612 (e.g., spectrum 1) can be mapped to multiple peptides 614a, 614b, 614c (e.g., peptides A, B, and C, respectively). Each peptide can be mapped to multiple accession numbers. For example, peptide 614a (e.g., peptide A) can be mapped to multiple accession numbers 616a, 616b, 616c, 616d, and 620 (e.g., accession numbers A, B, C, D, and E, respectively); peptide 614b (e.g., peptide B) can be mapped to accession numbers 620 and 624b (e.g., accession numbers E and G), and peptide 616c (e.g., peptide C) can be mapped to multiple accession numbers 624a, 624b, 624c, and 624d (e.g., accession numbers F, G, H, and I, respectively). Accession numbers can be mapped to different gene symbols. For example, accession number 624a (e.g., accession number F) can be mapped to gene 626 (e.g., gene C); accession number 624b (e.g., accession number G) can be mapped to gene 628 (e.g., gene D); accession number 624c (e.g., accession number H) can be mapped to gene 630 (e.g., gene E); and accession number 624d (e.g., accession number I) can be mapped to gene 632 (e.g., gene F). Accession numbers can also be mapped to the same gene symbol. For example, accession numbers 616a, 616b, 616c, and 616d (e.g., accession numbers F, G, H, and I, respectively) can be mapped to gene 618 (e.g., gene B). As can be seen from the example of Figure 6, the relationship between peptides, accession numbers, and gene symbols can result in highly redundant data.

[0161] One or more strategies can be employed to reduce the redundancy of the final data without losing information at the peptide, accession number, and gene symbol level.

[0162] For example, a strategy may include assigning a unique identifier (e.g., spectra ID) to each sample or MS / MS spectrum and selecting a best fit spectrum (BFPS) for each sample or MS / MS spectrum. The best fit can be determined by a likelihood metric. In some embodiments, a machine learning model can be used to determine the best fit.

[0163] Candidate peptide sequences can be generated by a combination of methods, including multiple libraries and multiple random generation methods. Each combination can be assigned a unique identifier (e.g., Search definition ID). There can be different methods to calculate the best fit. In some embodiments, the best fit can be calculated by Search definition ID, where for each Search definition ID, the best fit for each spectrum is selected independently of other Search definition IDs.

[0164] In some embodiments, the library and search engine can select the best fit (BFPS) of each spectrum. In the case of a library engine combination with multiple static search methods, the number can be split into the method with the highest spectra ID score or the highest likelihood index. In this case, the overlap between the two methods is minimal, unless the spectra ID scores equally high in both methods. Once the best fit of each spectrum is determined, a non-redundant list can be made for each level that can analyze data.

[0165] At the gene symbol level, only unique spectra ID-Search definition ID-Gene symbol combinations will be used to create quantities for all subgroups of the experiment and for further statistical analyses where groups are compared by gene symbol.

[0166] The same operation was done at the peptide level, where only the spectra ID-Search definition ID-peptide combination was brought into the quantity table and analyzed at the peptide level. The Search definition ID corresponds to the library and search settings.

[0167] The peptide identification system 110 was tested using angiotensin and GluFib samples, 18 standard protein samples, and plasma samples. The results were compared with existing systems (e.g., XITandem, (Proteome Discoverer)) were compared.

[0168] For the angiotensin and glutamate fibrinogen (GluFib) tests, single peptide standards of angiotensin and GluFib were prepared. A dilution series of 0.1, 5, 10, 50, 100, 500, 1000 μg / mL of GluFib and angiotensin was applied to the LTQ linear ion trap by direct injection. That is, the two peptide standards, angiotensin and GluFib, were injected into the mass spectrometer at increasing concentrations from 0 μg / mL to 1000 μg / mL, resulting in 16 MS runs (including replicates) for each standard. The peptide sequences of the two standards were DRVYIHPF for angiotensin and EGVNDNEEGFFSAR for GluFib.

[0169] The peptide recognition system 110 was able to recognize angiotensin and GluFib peptides, with the number of recognitions increasing as the concentration increased. For the random generation method, the peptide recognition system 110 was given constraints of 523 and 1046 Daltons, and for demonstration purposes, the peptide constraint was D*VYIHPF. Based on these constraints, the system generated four candidate peptides: DFVYIHPF, DRVYIHPF, DUVYIHPF, and DYVYIHPF, and subsequently only the angiotensin standard peptide DRVYIHPF was proposed.

[0170] For the library selection method, the peptide identification system 110 uses two peptide libraries for each sample. For the angiotensin sample test, the first library only stores angiotensin peptides, and the second library contains a total of 200 peptides, storing angiotensin peptides and hundreds of other peptide sequences from humans and other species, where the masses of the peptides are very close to the known masses of the sample angiotensin peptides at different charge states. Similarly, for the GluFib sample test, the first library only stores GluFib peptides, and the second library contains a total of 232 peptides, storing GluFib peptides and hundreds of other peptide sequences from humans and other species, where the masses of the peptides are very close to the known masses of the sample GluFib peptides at different charge states.

[0171] The peptide identification system 110 identified the angiotensin peptide DRVYIHPF638 times in all angiotensin samples using the candidate peptide sequences selected from the memory, and the selected number is displayed together with its score or likelihood indicator, as shown in FIG. Fig. 7A 700. Analysis of the spectrum of a high concentration of angiotensin peptide indicated that 114 charge state 1 recognitions and 24 charge state 2 recognitions should be expected. The results indicated 114 charge state 1 recognitions and 14 charge state 2 recognitions.

[0172] Similarly, direct spectrum analysis of the GluFib peptide indicated that it was expected that there would be 76 identifications of charge state 1 peptides and 76 identifications of charge state 2. The results showed that there were 76 identifications of charge state 1 and 75 identifications of charge state 2d. Therefore, the systems and methods disclosed in the present application correctly identified angiotensin and GluFib peptides.

[0173] In contrast, the existing systems XITandem and It performs poorly when set to search without enzyme. Specifically, XITandem returns no identity during the run, while Returns 1 GluFib identifier. It is worth noting that XITandem and Searches were not performed against limited libraries containing only angiotensin or only GluFib and other peptide sequences of similar mass.

[0174] The peptide identification system 110 uses the candidate peptide sequences selected from the memory and identifies the GluFib peptide EGVNDNEEGFFSAR868 times in all angiotensin samples. The selected number is displayed together with its score, as shown in Figure 7B In addition, the GluFib peptide EGVNDNEEGFFSAR is the highest scoring peptide and the third highest scoring peptide after taking the charge state into account, as shown in Table 720. Figure 7C As shown in Table 740.

[0175] A panel of 18 standard peptides was applied to nano- and micro-electrospray as described above for the 18 standard protein tests. Each sample was injected twice and was from 9 different species.

[0176] The 18 standard proteins come from 9 species. Therefore, for the library selection method, a protein library is assembled for each species, which contains only the protein standards belonging to that species. A small library containing the 18 proteins that the protein standards are expected to contain is also generated. All samples are tested against the small 18-protein library and the corresponding species library for that sample.

[0177] Since the library of 18 proteins was small and the number of digested peptides was small, there was limited competition between peptides for high-quality identification of specific MSn lines.

[0178] Now refer to Fig. 8A , Figure 8B and Figure 8C , wherein Figures 800, 810 and 820 respectively show the peptide recognition system 110, XITandem and Results of the 18 standard protein tests of (Proteome Discoverer). In graphs 800, 810, and 820, the vertical axis shows the number of identifications and the horizontal axis shows the gene symbols listed in alphabetical order. The black bars represent identifications where the candidate peptide sequence matches the sample, while the red lines represent identifications where the candidate peptide sequence does not match the sample.

[0179] from Fig. 8A , Figure 8B and Figure 8C It can be seen that the peptide recognition system 110 and the existing system performed well in the 18 standard protein tests. In addition, the peptide recognition system 110 was able to make accurate recognitions that were comparable to the existing system.

[0180] Searching in the all-species library tests the system's ability to attribute each scan to the correct theoretical peptide in the presence of more intense competition. Fig. 9A , Fig. 9B and Fig. 9C Figures 900, 910 and 920 respectively show the Fig. 8A , Figure 8B and Figure 8C The same set of runs shown in Figures 800, 810, and 820, respectively, are for a typical b. Taurus library containing 160,297 different proteins, which produced 5,285,355 peptides after in silico digestion with no missed cleavages. It can be seen that the results of the peptide identification system 110 when used with a large protein library are qualitatively (but not quantitatively) similar to those of existing systems.

[0181] Table 1 shows the number of different peptides returned by the system for peptides that are actually present in the reference protein library, that have not been missed during the in silico digestion, and for which the reported protein accession number matches the accession number in the library. Based on the number of precise identifications returned by each system, the peptide identification system 110 can more accurately identify the presence of a particular peptide. Table 1

[0182] For plasma testing, a total of 12 plasma 25 μL samples were sterilized with 250 μL 100% acetonitrile (HPLC grade) and centrifuged at 14,800 ref for 5 minutes. The supernatant was removed and the samples were freeze-dried overnight. The freeze-dried plasma samples were resuspended in 200 μL 20mM Tris, pH 8.8 loading buffer, applied to a QA column containing 100 μL resin, washed in 3 volumes of loading buffer, and eluted in 200 μL 300mM NaCl (20mM Tris pH 8.8). The resin was collected in 200 μL loading buffer. The QA salt extraction fraction was divided into 3 subgroups, 4 samples per subgroup, and digested by any of the following methods: 1. Trypsin digestion and partial reduction, 2. Trypsin digestion and reduction and alkylation, 3. Trypsin digestion without reduction or alkylation. The samples were applied to nano-LC-MSMS.

[0183] Now refer to Fig. 10A , Fig. 10B and Fig. 10C , wherein graphs 1000 , 1010 , and 1020 , respectively, are shown showing the results of the plasma test, i.e., the gene symbols identified by the peptide identification system 110 and the prior art system for the trypsin human library. Fig. 10A Overlapping gene symbols are displayed, with at least 1 peptide number and no missed cleavage; Fig. 10B Overlapping gene symbols are shown, with at least 3 peptide numbers and no missed cleavages; Fig. 10C Overlapping gene symbols are shown, with at least 10 peptide numbers and no missed cleavages.

[0184] The peptide identification system 110 can use two theoretical ion sources - a random generator or a known theoretical spectral library. The dual nature of the peptide identification system 110 can be advantageously exploited in a variety of ways. First, the spectral library is used to identify the best matching peptide sequence, and then the random generator is used to randomly generate combinations of the same amino acid sequence plus possible modifications until the best fit is obtained.

[0185] Second, when one source of theoretical spectra is used to fill in some amino acid gaps under certain criteria, the peptide identification system 110 can automatically switch to another source of theoretical spectra given certain thresholds or parameters. Third, a synthetic peptide library can be generated to include known amino acids at known positions, or, for example, based on partial sequences of known binding target binding sites, which reduces the number of peptides that may be randomly generated. Fourth, peptide sequences can be excluded from the random generator possibility pool. In addition, parameters can be set to change the random generator strategy based on certain events, thresholds, or parameters.

[0186] Now refer to Fig.11A, which shows a diagram of a linear regression model for fitting a query spectrum to various candidate peptide sequences according to an example embodiment. The fragment ions of the spectrum are fitted to theoretical MS / MS ions of five candidate peptide sequences 1100, 1110, 1120, 1130, 1140, and 1150. Fig.11A In the figure, the actual and theoretical m / z values ​​of each candidate peptide sequence are shown, and the size of each dot represents the difference between the actual mass and the theoretical mass. Fig. 11B A table 1160 is shown, which shows Fig.11A The data of the linear regression model for each candidate peptide sequence in (i.e., the first row contains results corresponding to candidate peptide sequence 1100, the second row corresponds to candidate peptide sequence 1110, the third row corresponds to candidate peptide sequence 1120, the fourth row corresponds to candidate peptide sequence 1140 and the fifth row corresponds to peptide sequence 1150). Based on these fits, one candidate peptide sequence (EGVNDNEEGFFSAR) can be determined as the best fit.

[0187] As mentioned above, various regression models can be used to fit the query spectrum. Figures 12A-12B , which shows Tables 1210 and 1220, showing the results of linear regression, chi-square, and cross-correlation regression models, respectively, for the GluFib peptide EGVNDNEEGFFSAR, searched against a small dairy cow library and reduced to the top 30 gene symbols. Table 1210 shows the results when the following filters are applied: YYZ_SequentialB>2AND YYZ_SequentialY>2AND log(lonMatchIntensityRatio)>2.2. Table 1220 shows the results when no filters are applied.

[0188] In some embodiments, the peptide identification system 110 can be operated to identify the best matching peptide sequence, and then another round is performed in which combinations of the same amino acid sequence plus possible modifications are considered until the best fit is obtained.

[0189] Finally, one can directly test whether the observed frequency of a peptide or aptamer sequence is simply a random fit by determining how often the sequence is obtained from random MS / MS or MSn spectra from in silico random or physical noise MS / MS or MSn spectra. The observed frequency can be calculated from the total number of MS / MS or MSn spectra collected or generated, or from the total number of fits of MS / MS or MSn spectra to real or imaginary peptide sequences.

[0190] It should be understood that many specific details are set forth in order to provide a thorough understanding of the example embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein can be implemented without these specific details. In other cases, well-known methods, procedures, and components are not described in detail to avoid blurring the embodiments described herein. In addition, this specification and the accompanying drawings should not be construed as limiting the scope of the embodiments described herein in any way, but should be construed as merely describing the implementation of the various embodiments described herein.

[0191] It should be noted that the degree terms used in this application, such as "substantially", "approximately" and "approximately", refer to a reasonable amount of deviation of the modified term so that the end result will not be significantly changed. These degree terms should be interpreted as including the deviation of the modified term, provided that such deviation does not negate the meaning of the modified term.

[0192] In addition, the wording "and / or" used in this application is intended to mean inclusive or. That is, "X and / or Y" is intended to mean X or Y or both. As a further example, "X, Y and / or Z" is intended to mean X or Y or Z or any combination thereof.

[0193] It should be noted that the term “coupled” used in the present application means that two elements can be coupled to each other directly or coupled to each other through one or more intermediate elements.

[0194] Embodiments of the systems and methods described herein may be implemented in hardware or software or a combination of both. These embodiments may be implemented as a computer program executed on a programmable computer, each computer including at least one processor, a data storage system (including volatile memory or non-volatile memory or other data storage elements or a combination thereof), and at least one communication interface. For example, but not limited to, a programmable computer (hereinafter referred to as a computing device) may be a server, a network device, an embedded device, a computer expansion module, a personal computer, a laptop computer, a personal data assistant, a cellular phone, a smart phone device, a tablet computer, a wireless device, or any other computing device that can be configured to perform the methods described herein.

[0195] In some embodiments, the communication interface can be a network communication interface. In the embodiment of the combination element, the communication interface can be a software communication interface, such as a communication interface for interprocess communication (IPC). In other embodiments, the communication interface can be combined into a combination of hardware, software and hardware and software.

[0196] Program code may be applied to input data to perform the functions described herein and generate output information. The output information is applied to one or more output devices in a known manner.

[0197] Each program can be implemented in a high-level program or object-oriented programming and / or scripting language or both to communicate with a computer system. However, if necessary, the program can be implemented in assembly language or machine language. In any case, the language can be a compiled language or an interpreted language. Each such computer program can be stored on a storage medium or device (e.g., ROM, disk, CD) readable by a general or special programmable computer for configuring and operating the computer when the computer reads the storage medium or device to execute the program described in this application. The embodiment of the system can also be considered to be implemented as a non-transient computer-readable storage medium, configured with a computer program, wherein the storage medium so configured causes the computer to operate in a specific and predefined manner to perform the functions described in this application.

[0198] In addition, the systems, processes, and methods of the embodiments can be distributed in a computer program product that includes a computer-readable medium carrying computer-usable instructions for one or more processors. The medium can be provided in various forms, including one or more floppy disks, optical disks, tapes, chips, wired transmissions, satellite transmissions, Internet transmissions or downloads, magnetic and electronic storage media, digital and analog signals, etc. The computer-usable instructions can also take various forms, including compiled and non-compiled code.

[0199] This application describes various embodiments by way of example only. Various modifications and changes may be made to these exemplary embodiments without departing from the spirit and scope of the present invention, which is limited only by the appended claims.

Claims

1. A method for identifying peptides in a query spectrum, the method comprising: receiving one or more parameters of the query profile; generating one or more candidate peptide sequences based on the one or more parameters of the query spectrum; generating a plurality of samples of the query spectrum; selecting at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences; determining a likelihood indicator for each of the one or more candidate peptide sequences based on the comparison with the at least one sample; applying a signal-to-noise filter to the one or more candidate peptide sequences based on the likelihood indicators of the candidate peptide sequences; as well as Based on the filtered candidate peptide sequences, at least one candidate peptide sequence is selected as a proposed peptide sequence in the query spectrum.

2. The method according to claim 1, characterized in that Selecting at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences, for each sample, comprising: counting the number of consecutive b and y lines that match the sample; and If the number of the consecutive b and y lines matching the sample is less than a predetermined minimum number of lines, the sample is excluded from comparison with the one or more candidate peptide sequences.

3. The method according to any one of claims 1 to 2, characterized in that Selecting at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences, for each sample, comprising: determining a spectral intensity of the sample; Determine the number of matching b and y lines; For the matched b and y spectral lines, determining the sum of the corresponding spectral intensities; determining a sum of the spectral intensities; determining a ratio of the sum of the spectral intensities of the matched b and y spectral lines to the total of the spectral intensities; and If the ratio is less than a predetermined threshold, the sample is excluded from comparison with the one or more candidate peptide sequences.

4. The method according to any one of claims 1 to 3, characterized in that Selecting at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences, for each sample, comprising: determining a precursor mass of the sample; and If the precursor mass is substantially equal to the mass of the candidate peptide sequence, the sample is selected for comparison with the one or more candidate peptide sequences.

5. The method according to claim 4, characterized in that The precursor masses include at least one of the precursor masses having a charge state of 1, 2, or 3.

6. The method according to claim 4, characterized in that The precursor mass includes mass shifts from one or more post-translational modifications.

7. The method according to any one of claims 1 to 6, characterized in that Selecting at least one sample from the plurality of samples for comparison with the one or more candidate peptide sequences, for each sample, comprises: determining the mass of theoretical ions of the sample; and If the mass of the theoretical ion is substantially equal to the mass of the base peak, the sample is selected for comparison with the one or more candidate peptide sequences.

8. The method according to any one of claims 1 to 7, characterized in that Based on the comparison with the at least one sample, a likelihood index for each of the one or more candidate peptide sequences is determined, the likelihood index comprising using one or more of linear regression, multivariate linear regression, nested regression, or a linear model.

9. The method according to any one of claims 1 to 8, characterized in that Determining a likelihood index of each of the one or more candidate peptide sequences based on a comparison with the at least one sample, for each sample, comprising: The theoretical ions of the sample were compared with the corresponding theoretical ions of the candidate peptide sequence using a chi-square test.

10. The method according to any one of claims 1 to 9, characterized in that Determining a likelihood index of each of the one or more candidate peptide sequences based on a comparison with the at least one sample, for each sample, comprising: A cross-correlation score is determined relative to the candidate peptide sequence.

11. The method according to any one of claims 1 to 10, characterized in that Applying a signal-to-noise filter to the one or more candidate peptide sequences based on the likelihood indicators of the candidate peptide sequences comprises: Generate high frequency random source noise spectrum; determining a difference between the high frequency random source noise spectrum and the one or more candidate peptide sequences; and if the difference exceeds a predetermined threshold difference, excluding the sample as a selection of a proposed peptide sequence for the query spectrum; Otherwise, the sample is included as a selection for the proposed peptide sequence.

12. The method according to any one of claims 1 to 11, characterized in that Generating the high frequency random source noise spectrum includes using Monte Carlo stochastic simulation.

13. The method according to any one of claims 1 to 12, characterized in that Selecting the at least one candidate peptide sequence as a proposed peptide sequence for the query spectrum includes selecting the best fit for each spectrum at one of a peptide, accession number, or gene symbol level.

14. The method according to any one of claims 1 to 13, characterized in that Generating the one or more candidate peptide sequences based on the one or more parameters of the peptide query comprises one or more of the following: storing a plurality of peptide sequences in a computer readable medium; selecting at least one of the plurality of peptide sequences stored in the computer-readable medium for use as the at least one candidate peptide sequence based on the one or more parameters; and The at least one candidate peptide sequence is randomly generated based on the one or more parameters.

15. The method according to claim 14, characterized in that: Receiving one or more parameters of the query profile includes: receiving the one or more parameters from a user input at a computing device; The one or more parameters include one or more of an amino acid, a position of an amino acid, a group of amino acids, or a position of a group of amino acids.

16. The method according to any one of claims 14 or 15, characterized in that: Receiving one or more parameters of the query spectrum includes: receiving the query spectrum; and deriving the one or more parameters from the query spectrum; The one or more parameters include one or more of neutral loss, post-translational modification shift, iminium ions, subtraction of B ions, or subtraction of Y ions.

17. The method according to any one of claims 14 to 16, characterized in that Selecting at least one peptide sequence from the plurality of peptide sequences stored in the computer-readable medium based on the one or more parameters to serve as the at least one candidate peptide sequence comprises: identifying a subset of peptide sequences from the stored plurality of peptide sequences, each of the peptide sequences satisfying the one or more parameters; and The at least one candidate peptide sequence is selected from the subset of peptide sequences.

18. The method according to any one of claims 14 to 17, characterized in that The stored plurality of peptide sequences include one or more of the following: naturally occurring peptide sequences and synthetic peptide sequences.

19. The method according to any one of claims 14 to 18, characterized in that Randomly generating the at least one candidate peptide sequence based on the one or more parameters comprises: randomly generating the peptide sequence; determining whether the randomly generated peptide sequence satisfies the one or more parameters; and If the randomly generated peptide sequence meets the one or more parameters, the randomly generated peptide sequence is used as the candidate peptide sequence, otherwise the randomly generated peptide sequence is discarded.

20. The method according to any one of claims 14 to 19, characterized in that Each of the at least one randomly generated candidate peptide sequence has a predetermined length.

21. The method according to claim 19, characterized in that The at least one candidate peptide sequence is randomly generated based on the one or more parameters, and for each randomly generated candidate peptide sequence, the method comprises: assigning at least one of the amino acids to the randomly generated candidate peptide sequence based on the one or more parameters; randomly selecting an unallocated position of the predetermined length; For the unallocated locations, including: randomly selecting the amino acids; and assigning the amino acid to the unassigned position; and Continue to randomly assign the amino acids to the randomly generated candidate peptide sequence until every position has been assigned.

22. The method according to claim 21, characterized in that Randomly selected amino acids include: randomly selecting an amino acid group from a plurality of said amino acid groups, each of said amino acid groups comprising a plurality of said amino acids; and The amino acid is randomly selected from the group of randomly selected amino acids.

23. The method according to claim 22, characterized in that Each of the randomly generated candidate peptide sequences comprises at least one of the amino acids from each of the plurality of amino acid groups.

24. The method according to any one of claims 1 to 23, characterized in that Generating a plurality of samples of the query spectrum includes one or more of generating experimental spectra or generating simulated spectra.

25. The method according to claim 24, characterized in that Generating the simulated spectrum includes using the Monte Carlo stochastic simulation.

26. A system for identifying peptides of a query spectrum, the system comprising a non-transitory computer readable medium and a processor operable to: receiving one or more parameters of a query spectrum; generating one or more candidate peptide sequences based on the one or more parameters of the query spectrum; generating a plurality of samples of the query spectrum; selecting at least one sample from the plurality of samples to compare with the one or more candidate peptide sequences; determining a likelihood indicator for each of the one or more candidate peptide sequences based on the comparison with the at least one sample; applying a signal-to-noise filter to the one or more candidate peptide sequences based on the likelihood indicators of the candidate peptide sequences; as well as The at least one candidate peptide sequence is selected as a proposed peptide sequence for the query spectrum based on the filtered candidate peptide sequences.

27. The system according to claim 26, characterized in that For each sample, the processor is operable to: counting the number of consecutive b and y lines matching the sample; and If the number of consecutive b and y lines matching the sample is less than a predetermined minimum number of lines, the sample is excluded from comparison with the one or more candidate peptide sequences.

28. The system according to claim 27, characterized in that For each sample, the processor is operable to: determining a spectral intensity of the sample; For matched consecutive b and y spectral lines, determine the sum of the corresponding spectral intensities; Determine the sum of the spectral intensities; determining a ratio of a sum of spectral intensities of matched consecutive b and y spectral lines to a sum of said spectral intensities; and If the ratio is less than a predetermined threshold, the sample is excluded from comparison with the one or more candidate peptide sequences.

29. A system according to any one of claims 26 to 28, characterised in that For each sample, the processor is operable to: determining a precursor mass of the sample; and If the precursor mass is substantially equal to the mass of a candidate peptide sequence, the sample is selected for comparison with the one or more candidate peptide sequences.

30. The system according to claim 29, characterized in that The precursor masses include at least one of the precursor masses having a charge state of 1, 2, or 3.

31. The system according to claim 29, characterized in that The precursor mass includes mass shifts from one or more post-translational modifications.

32. A system according to any one of claims 26 to 31, characterized in that For each sample, the processor is operable to: determining the mass of theoretical ions of the sample; and If the mass of the theoretical ion is substantially equal to the mass of the base peak, the sample is selected for comparison with the one or more candidate peptide sequences.

33. A system according to any one of claims 26 to 32, characterized in that For each sample, the processor can be operable to determine a likelihood indicator for the sample based on comparison with the one or more candidate peptide sequences using one or more of linear regression, multiple linear regression, or nested regression.

34. A system according to any one of claims 26 to 33, characterized in that For each sample, the processor is operable to compare theoretical ions of the sample with corresponding theoretical ions of the candidate peptide sequence using a chi-square test to determine a likelihood index for the sample.

35. A system according to any one of claims 26 to 34, characterised in that For each sample, the processor is operable to determine a cross-correlation score relative to the candidate peptide sequence to determine a likelihood indicator for each of the at least one sample based on comparison with the one or more candidate peptide sequences.

36. A system according to any one of claims 26 to 35, characterised in that The processor is operable to: Generate high frequency random source noise spectrum; determining a difference between the high frequency random source noise spectrum and the one or more candidate peptide sequences; and If the difference exceeds a predetermined threshold difference, excluding the sample as a selection of a proposed peptide sequence for the query spectrum; Otherwise, the sample is included as a selection for the proposed peptide sequence.

37. The system according to claim 36, characterized in that The processor can generate the high frequency random source noise spectrum using Monte Carlo stochastic simulation.

38. A system according to any one of claims 26 to 37, characterized in that The processor is operable to select the best fit for each profile at one of the level of peptide, accession number or gene symbol.

39. A system according to any one of claims 26 to 38, characterized in that The processor is operable to: storing a plurality of peptide sequences in a computer readable medium; selecting at least one of the plurality of peptide sequences stored in the computer-readable medium for use as at least one candidate peptide sequence of the one or more candidate peptide sequences based on the one or more parameters; and At least one candidate peptide sequence among the one or more candidate peptide sequences is randomly generated based on the one or more parameters.

40. The system according to claim 39, characterized in that The processor is operable to: The one or more parameters are received from a user input of a computing device, the one or more parameters comprising one or more of an amino acid, a position of an amino acid, a group of amino acids, or a position of a group of amino acids.

41. A system according to any one of claims 39 or 40, characterized in that The processor is operable to: receiving the query spectrum; and deriving the one or more parameters from the query spectrum; The one or more parameters include one or more of neutral loss, post-translational modification shift, iminium ions, subtraction of B ions, or subtraction of Y ions.

42. A system according to any one of claims 39 to 41, characterised in that The processor is operable to: identifying a subset of peptide sequences from the stored plurality of peptide sequences, each of the peptide sequences satisfying the one or more parameters; and At least one candidate peptide sequence is selected from the subset of peptide sequences.

43. A system according to any one of claims 39 to 42, characterised in that The plurality of stored peptide sequences include one or more of the following: naturally occurring peptide sequences and synthetic peptide sequences.

44. A system according to any one of claims 39 to 43, characterised in that The processor is operable to: randomly generating the peptide sequence; determining whether the randomly generated peptide sequence satisfies the one or more parameters; and If the randomly generated peptide sequence meets the one or more parameters, the randomly generated peptide sequence is used as the candidate peptide sequence, otherwise the randomly generated peptide sequence is discarded.

45. A system according to any one of claims 39 to 44, characterised in that Each of the at least one randomly generated candidate peptide sequence has a predetermined length.

46. ​​The system according to claim 45, characterized in that The processor is operable to: assigning at least one of the amino acids to the randomly generated candidate peptide sequence based on the one or more parameters; randomly selecting an unallocated position of the predetermined length; For the unassigned locations: randomly selecting the amino acids; and assigning the amino acid to the unassigned position; and Continue to randomly assign amino acids to the randomly generated candidate peptide sequence until every position has been assigned.

47. The system according to claim 46, characterized in that The processor is operable to: randomly selecting an amino acid group from a plurality of said amino acid groups, each of said amino acid groups comprising a plurality of said amino acids; and The amino acid is randomly selected from the group of randomly selected amino acids.

48. The system according to claim 47, characterized in that Each of the randomly generated candidate peptide sequences comprises at least one of the amino acids from each of the plurality of amino acid groups.

49. A system according to any one of claims 26 to 48, characterized in that The processor is operable to generate an experimental spectrum or a simulated spectrum.

50. The system according to claim 49, characterized in that The processor can generate the simulated spectrum using the Monte Carlo stochastic simulation.

Images