JP2017096668A - Identification support method and identification support device for living matter derived substance - Google Patents

Abstract

PROBLEM TO BE SOLVED: To support a work for identifying protein, and improve identification accuracy.SOLUTION: An identification support method is configured to clearly indicate and display a portion corresponding to peptide which is estimated in database retrieval in an amino acid sequence for every protein candidate, when database retrieval based on a mass spectrum acquired by mass analysis, is performed, plural candidates whose scores are close, are obtained, namely a result is not determined (S9). If plural mass spectra are included in one analysis unit, determined peptide and non-determined peptide obtained from the plural mass spectra are clearly indicated and displayed in the amino acid sequence of the protein candidate. An analyzer looks at the display, then selects a proper candidate by determining possibility of attribution such as, the determined peptide is included (S10). Then, through a manual attribution of the analyzer, a final identification result is determined and displayed (S11).SELECTED DRAWING: Figure 2

Description

  The present invention relates to a method for supporting the work of mass-analyzing a test sample containing a biological substance and identifying the biological substance using mass spectrometry data obtained thereby, and an apparatus for performing the support. . In particular, the present invention mass-analyzes a test sample containing a peptide mixture derived from a protein that is one of biological substances, and estimates the amino acid sequence of the peptide using the mass spectrometry data obtained thereby. The present invention relates to an identification support method and apparatus suitable for supporting an identifying operation.

In recent years, protein structures and functions have been rapidly analyzed as post-genomic research. As one of such protein structure / function analysis methods (proteomics), protein expression analysis and primary structure analysis using mass spectrometers have been widely performed. So-called MS ⁿ analysis (n is an integer of 2 or more), which captures and cleaves specific peaks by collision-induced dissociation (CID) or the like, is effective. In general, in MS ² analysis, ions having a specific mass-to-charge ratio are first selected as precursor ions from an analysis target, and the precursor ions are cleaved by CID or the like. Thereafter, by analyzing the mass of various product ions generated by cleavage, information on the mass and chemical structure of the target ion can be obtained.

When identifying a protein using MS ⁿ analysis, first, the protein to be identified is digested with an appropriate enzyme to form a mixture of peptide fragments, and then mass spectrometry is performed on the peptide mixture to obtain a mass spectrum. . Subsequently, from a mass spectrum data of the peptide mixture, for example, a set of isotope peaks derived from a single peptide is selected as a precursor ion, the precursor ion is cleaved by CID or the like, and the generated product ion Mass spectrometry for MS, ie MS ² analysis. In addition, when the cleavage operation is not performed to a sufficiently small fragment by one cleavage operation, the cleavage operation is performed a plurality of times, that is, MS ⁿ analysis in which n is 3 or more may be performed.

If an MS ⁿ spectrum is obtained by mass spectrometry with one or two or more multi-stage cleavage operations as described above, for example, the mass of the product ion obtained from the MS ⁿ spectrum and the protein ion database are registered. The mass of the peptide fragment obtained by calculation from the amino acid sequence of the protein in question is compared, the amino acid sequences of a plurality of test peptides are determined using the degree of coincidence as a clue, and the target protein is identified.
As analysis software for performing such a database search by a computer, MS / MS ion search included in Mascot of Matrix Science, UK is well known. The same function is also installed in other analysis software such as X! Tandem provided by GPM (The Global Proteome Machine Organization).

In a conventional protein identification system using MS / MS ion search or the like, not only the corresponding peptide but also the protein is automatically identified from the peak appearing in the MS ⁿ spectrum. However, multiple candidate peptides that can be assigned from one MS ⁿ spectrum may be obtained. Even if the peptide is assigned to one peptide, the peptide appears not only in a single protein but also in multiple types of proteins. There is also a guarantee that the protein can be automatically determined uniquely because, for example, the post-translational modification may cause the peptide mass-to-charge ratio to differ from that of the amino acid sequence recorded in the database. Absent. In particular, in identification work aiming at proteomics research, the test sample is often in a state in which peptides derived from a plurality of types of proteins are mixed, not a peptide mixture derived from a single protein. In such a case, the reliability of the automatically derived protein identification result is not necessarily high, and it is necessary for the analyst to judge the reliability.

  Therefore, in the analysis using the software such as Mascot described above, the candidate peptides that are estimated to be applicable based on the mass-to-charge ratio of the product ions, together with the entry names of the proteins that each include the amino acid sequence, A score, which is an index indicating the degree of matching in database search, is output for each candidate. In addition, when using the protein view (Protein View) function (refer to Non-Patent Document 1) installed in Mascot, the peptide assigned by database search is located at any position in the amino acid sequence of the assigned protein. (For example, the amino acid sequence of the assigned peptide is displayed in red and the others are displayed in black). Furthermore, the proportion of the identified peptide in the amino acid sequence of the entire protein is displayed by an index called protein sequence coverage. Based on such a display, the analyst can determine how reliable the result of identifying the protein in the analysis is.

  In the conventional protein identification system, any of the above-mentioned information provided to the analyst assigns the estimated peptide as a result of the database search to the protein, and then determines the reliability of the assignment to the protein. It is evaluated after execution of automatic identification processing. Therefore, if the analyst determines that the reliability of the protein identification result is low, for example, the peptide estimation result obtained in the process of identifying the protein is verified, and the selection at the stage of assigning the peptide to the protein is performed. The analyst himself needs to start over. One index referred to in such a case is a score given to each candidate peptide obtained as an identification result, but in fact, even the highest score among the scores given to multiple candidates, Often, a statistically significant score has not been reached. In addition, a plurality of identification candidates are derived, and even if all of them are significant, the score is not high enough, and the score values are often close to each other. It can be said that such a state is a state in which at least a part of the assignment result is not uniquely determined. In such a case, it is not easy to select a protein candidate considered to be most appropriate.

In other words, in the conventional protein identification system, even if the analyst tries to perform protein identification (assignment) based on his / her own judgment using the peptide identification result based on the MS ⁿ spectrum, the analyst makes such judgment. Are not provided with sufficient materials. Therefore, the task of manually identifying a protein by an analyst is not only very troublesome, but it is also difficult to increase the accuracy of the identification.

  In addition, in the conventional protein identification system using the analysis software described above, a plurality of protein candidates extracted by the identification work are displayed in a ranking format, and many analysts indicate that the first ranking protein in the displayed ranking. Candidates are adopted as they are, and results below the second rank are not considered. However, when there are a plurality of similar amino acid sequences such as isoforms, the possibility that a more appropriate result is included in the candidates of the second rank or lower cannot be denied. When the post-translational modification is included in the amino acid sequence to be analyzed, especially when a plurality of protein candidates having different amino acid sequences of the peptide and different post-translational modifications are obtained, the ranking is first. It is not easy to determine whether a candidate is appropriate. At present, in such a case, there is not yet a determination algorithm or an evaluation function that is reliable enough to allow the computer to completely decide which attribution is most appropriate to select. Therefore, although it is desirable for the analyst to make the final determination of protein assignment, appropriate support methods for making such a determination easily, especially information useful for the analyst to make such a determination visually There is no method for displaying the information easily.

  Furthermore, in the analysis of biological materials, identification results (for example, a list of assigned proteins) obtained by database search as described above are obtained from other omics measurement data such as genomics and metabolomics. Often, omics analysis is performed that is integrated with data obtained by measurement and analysis such as transcriptomics. For example, in the identification of proteins using omics analysis, peptides are identified based on the results of mass spectrometry, and then the genome information obtained based on the same mass analysis results is used for the protein that is the basis of the peptides. May be automatically identified.

  However, in fact, as described above, there is no guarantee that peptides and proteins can be uniquely determined automatically in the process of assigning peptides and proteins using mass spectrometry results, so genomic information is used as described above. Even so, the protein cannot always be identified accurately.

  The same problem in omics analysis is that the cleavage pattern is not grasped more than the protein, genome information is not directly available, and the database for identification is not fully prepared. It also occurs when identifying other biological substances such as lipids and sugar chains.

“Mascot database search Result report overview” [online], Matrix Science Ltd., UK [May 15, 2015 search], Internet < URL: http://www.matrixscience.com/help/results_help.html#PROT>

  The present invention has been made in order to solve the above-mentioned problems. The first object of the present invention is to provide appropriate information to the analyst during the identification process, particularly when identifying proteins among biological materials. By providing, the identification assistance method and identification assistance apparatus of the biological substance which can perform the identification with the high precision which utilized the judgment of this analyst in part are provided.

  In addition, the second object of the present invention is to be able to improve the identification accuracy of biological substances by using other omics information such as use of genomics measurement results in proteomic analysis and other related information. The object is to provide a substance identification support method and an identification support apparatus.

The identification support method for a biological substance according to the first aspect of the present invention made to achieve the first object is a result of mass spectrometry of a sample containing a peptide mixture obtained by fragmenting a target protein that is a biological substance. An identification support method for a biological substance that supports identification work for identifying a target protein based on
a) Candidates that estimate the amino acid sequences of at least some of the peptides contained in the sample based on mass spectral data obtained by mass spectrometry of the sample and extract a plurality of protein candidates as target proteins based on the estimation result An extraction step;
b) For each of a plurality of protein candidates extracted in the candidate extraction step, the amino acid sequence of the peptide estimated in the candidate extraction step can be distinguished from other amino acid sequences in the description of the amino acid sequence of the protein. A display processing step to display as it is,
And the analyst can select a candidate from protein candidates by referring to the display result in the display processing step.

Moreover, the identification support apparatus for a biological substance according to the first aspect of the present invention, which has been made to achieve the first object, mass-analyzed a sample containing a peptide mixture obtained by fragmenting a target protein that is a biological substance. An identification support apparatus for a biological substance that supports identification work for identifying a target protein based on a result,
a) Candidates that estimate the amino acid sequences of at least some of the peptides contained in the sample based on mass spectral data obtained by mass spectrometry of the sample and extract a plurality of protein candidates as target proteins based on the estimation result An extractor;
b) For each of a plurality of protein candidates extracted by the candidate extraction unit, the amino acid sequence of the peptide estimated by the candidate extraction unit can be distinguished from other amino acid sequences in the description of the amino acid sequence of the protein. A display processing unit for displaying as it is,
The analyzer is capable of selecting a candidate from protein candidates by referring to the display result by the display processing unit.

Here, the mass spectrum data includes MS ⁿ spectrum data in which n is 2 or more.

  In the Mascot Protein View described above, peptides whose attribution was determined by database search based on mass spectrum data were highlighted in the notation of the amino acid sequence of the protein whose attribution was also automatically determined. On the other hand, in the identification support method according to the first aspect of the present invention realized by the identification support device according to the first aspect of the present invention, based on the stage before the protein assignment is determined, that is, based on the mass spectrum data. When a protein containing the estimated peptide is listed as a candidate for the target protein, the analyst is provided with a display that clearly indicates which part of the amino acid sequence of the protein candidate the estimated peptide corresponds to. For example, even if there is only one estimated peptide, generally, there are a plurality of proteins containing the one peptide, and therefore a plurality of protein candidates are extracted. For each of the plurality of protein candidates, for example, the amino acid sequence corresponding to the peptide is displayed in a different character color so that the estimated position of the peptide becomes clear. The analyst may look at this display and determine himself / herself which candidate is to be adopted, that is, perform an attribution determination manually.

For example, ^if there is one MS ⁿ spectrum to be analyzed, it is usually derived from one peptide, so there is only one peptide specified in the amino acid sequence of the protein candidate. On the other hand, if all of the peptides estimated based on multiple MS ⁿ spectra obtained from one sample that is a peptide mixture are clearly shown in the amino acid sequence of the protein candidate, this is clearly indicated for each protein candidate. More material is referred to when the analyst selects candidates such as the number of peptides varies. This not only facilitates the judgment by the analyst, but also increases the probability that an accurate selection can be made, that is, a correct protein is selected.

Therefore, in the identification support method for a biological substance according to the first aspect of the present invention, preferably,
An analysis unit specifying step in which an analyst specifies a range of data to be reflected in one analysis as one analysis unit;
In the candidate extraction step, the estimation of the amino acid sequence of the peptide and the extraction of protein candidates are performed based on all data designated as one analysis unit,
In the display processing step, peptide estimation results based on all data designated as one analysis unit may be integrated and displayed for each protein candidate.

Here, the designation of “data range to be reflected in one analysis” is, for example, data constituting one mass spectrum, data constituting all mass spectra obtained for one sample, It may mean selection of data or the like constituting all mass spectra obtained for a plurality of samples performed in one experiment. As one analysis unit, if data constituting all mass spectra obtained for one sample containing a peptide mixture derived from one target protein is selected, a plurality of amino acid sequences of protein candidates as described above are selected. Since a peptide may be clearly specified, an appropriate candidate can be selected using, for example, the number of peptides as one material.
On the other hand, if data constituting all mass spectra obtained for a plurality of samples carried out in one experiment is selected as one analysis unit, all data expressed under the conditions of the experiment Therefore, exhaustive, that is, proteomic analysis is performed.

  In a database search using MS / MS ion search or the like, a score is calculated when a peptide is estimated. If the score value is extremely high, the peptide is almost deterministic. Conversely, if the score is low, the peptide is likely but not deterministic. Therefore, when specifying peptides in the amino acid sequences of protein candidates, peptides that are determined to be deterministic and peptides that are determined to be non-deterministic are displayed in a distinguishable manner according to the score. May be. Moreover, it is good also as a display which can understand the range of the score value provided to each peptide visually.

The identification support method for a biological substance according to the second aspect of the present invention made to achieve the second object is based on the result of mass spectrometry of a sample containing the target biological substance. A method for supporting identification of a biological substance that supports identification work for identifying a substance,
a) a candidate extraction step of extracting a plurality of candidate substances of a target biological substance based on mass spectral data obtained by mass spectrometry on a sample;
b) a substance selection step of selecting a highly probable substance candidate from among the plurality of candidate substances using reference information related to the target biological substance that cannot be derived from the mass spectrum data;
It is characterized by having.

Moreover, the identification support apparatus for a biological substance according to the second aspect of the present invention, which has been made to achieve the first object, provides a target biological substance based on the result of mass spectrometry of a sample containing the target biological substance. An identification support apparatus for a biological substance that supports identification work for identifying a derived substance,
a) a candidate extraction unit that extracts a plurality of candidate substances of a target biological substance based on mass spectral data obtained by mass spectrometry of a sample;
b) using a reference information related to the target biological substance that cannot be derived from the mass spectrum data, a substance selection unit that selects a highly probable substance candidate from the plurality of candidate substances;
It is characterized by having.

  The term “biologically derived material” as used herein refers to proteins, nucleic acids, sugar chains, lipids, and the like. In the identification support method according to the second aspect of the present invention, for example, when the biological substance is a protein, the candidate extraction step uses a database search using MS / MS ion search or the like, or an amino acid sequence using de novo sequencing. By the analysis, a plurality of target protein candidates are extracted based on the mass spectrum data. In the conventional general identification work, a highly probable protein candidate is selected based on information such as the reliability of the peptide obtained in the process of estimating the amino acid sequence of the protein, that is, based on the result of mass spectrometry. On the other hand, in the identification support method according to the second aspect of the present invention, information that is not obtained by using the mass spectrometry result and reliable reference information is used to select a plurality of protein candidates. Select a candidate with a high probability.

  Information that can be used as the reference information includes information obtained by an omics analysis of a type different from the analysis method used for extracting candidate substances in the candidate extraction step. Specifically, when the target biological substance is a protein, and protein candidates are extracted by a mass spectrometry proteomic analysis method in the candidate extraction step, the above reference information other than the results of the mass spectrometry proteomic analysis Other omics information can be used. As such omics information, for example, information on gene expression level by genomics analysis, pathway map showing gene expression network in genomics analysis, pathway map showing metabolite network in metabolomics analysis, etc. are useful. It is also conceivable to use proteomic analysis results using, for example, antibodies other than those obtained by mass spectrometry as omics information.

  Such omics information is obtained by a method different from proteomics analysis by mass spectrometry, and is a result obtained by accumulating knowledge from the past, and thus has sufficient reliability. Thus, such omics information can be useful information in determining protein attribution based on mass spectrometry results.

  In addition, when the target biological substance is a protein and protein candidates are extracted by a proteomic analysis method using mass spectrometry in the candidate extraction step, the three-dimensional structure information of the protein, in addition to the omics information, is used as the reference information. It can also be used. In general, protein three-dimensional structure information is difficult to estimate from only the amino acid sequence estimated from the results of mass spectrometry, and is obtained based on, for example, X-ray structure analysis information obtained by a method completely different from mass spectrometry. It is

  According to the identification support method and the identification support apparatus for a biological substance according to the first aspect of the present invention, protein assignment is not automatically determined based on the mass spectrum, but is estimated based on the mass spectrum. The result of the analysis can be compared visually by the analyst, and the attribution of the protein can be determined based on the judgment of the analyst. For this reason, even if some of the attribution results are not uniquely determined and the reliability of automatic protein attribution is low, judgment based on the analyst's own knowledge and experience is added, and appropriate protein candidates are added. Can be easily selected. This eliminates the troublesome task of an analyst re-assigning again when the result of the protein assignment is low after the result of the protein assignment, and the accuracy of protein identification can be improved.

  According to the biological substance-derived substance identification support method and the identification support apparatus according to the second aspect of the present invention, in the estimation of the biological substance based only on the mass analysis result, the scores of a plurality of candidates are close to each other. Even when it is impossible to determine whether the correct answer is correct, the possibility of selecting a correct candidate is increased by using reference information such as omics information that already has a certain degree of reliability. In this way, the accuracy of identification of biological substances such as proteins can be improved.

BRIEF DESCRIPTION OF THE DRAWINGS The whole block diagram of the protein identification system which is 1st Example of this invention. The flowchart which shows the identification process procedure in the protein identification system of 1st Example. The figure which shows the example of the display output in the protein identification system of 1st Example. The whole block diagram of the protein identification system which is 2nd Example of this invention. The flowchart which shows the identification process procedure in the protein identification system of 2nd Example. Explanatory drawing of the identification example in the protein identification system of 2nd Example. Explanatory drawing of the identification example in the protein identification system of 2nd Example. Explanatory drawing of the identification example in the protein identification system of 2nd Example.

[First embodiment]
Hereinafter, a protein identification system according to a first embodiment of the present invention will be described with reference to the accompanying drawings.
FIG. 1 is an overall configuration diagram of the protein identification system of the first embodiment, and FIG. 2 is a flowchart showing an identification processing procedure in the protein identification system of the first embodiment.

  The protein identification system of the first embodiment includes a mass analysis unit 1, a data analysis unit 2, an input unit 3, and a display unit 4. The data analysis unit 2 includes a database search unit 21, a protein database 22, a search result extraction unit 23, a non-confirmed result storage unit 24, a confirmation result storage unit 25, a search result integration unit 26, a display processing unit 27, and a result confirmation processing unit 28. Are included as function blocks. Other than the mass analysis unit 1, it can be configured mainly by a personal computer or the like, and main functions are realized by executing dedicated software installed in the computer on the computer.

  The mass analyzer 1 is, for example, an ion trap time-of-flight mass spectrometer. Although not shown, the ion source that ionizes molecules and atoms in the target sample and the generated ions are temporarily captured, and the mass is obtained as necessary. A three-dimensional quadrupole ion trap that performs ion selection according to charge ratio and ion cleavage by CID, and various ions ejected from the ion trap are separated and detected according to mass-to-charge ratio. A time-of-flight mass spectrometer. As an ion source for performing ionization on a biological sample, one using a matrix-assisted laser desorption ionization (MALDI) method, one using an electrospray ionization (ESI) method, or the like is generally used.

  The database search unit 21 performs database search using an MS / MS ion search search engine installed in the existing analysis software such as Mascot described above. The protein database 22 contains the amino acid sequences of known proteins. For example, a publicly available database such as Swiss-Prot can be used.

An example of a protein identification procedure using the protein identification system of the first embodiment will be described with reference to FIG.
First, a peptide mixture prepared by enzymatic digestion of a target protein is used as a test sample, and MS ² analysis is performed in the mass spectrometer 1. Specifically, a mass spectrum is obtained by performing normal mass spectrometry without CID operation on one test sample, and a peak that is assumed to be derived from a peptide appearing in the mass spectrum is targeted. MS ² analysis is performed to obtain an MS ² spectrum. Usually, since a plurality of peptides-derived peaks appear in the mass spectrum for the test sample, MS ² analysis is performed on each of the peaks to obtain an MS ² spectrum. The data constituting the mass spectrum and the MS ² spectrum thus acquired are temporarily stored in the spectrum data storage unit 20. Further, when there are a plurality of test samples, a mass spectrum and an MS ² spectrum are acquired for each of the plurality of test samples, and data constituting them are temporarily stored in the spectrum data storage unit 20.

When performing protein identification based on the data collected in this way, the analyst first designates an analysis unit to be processed in one analysis from the input unit 3 (step S1). As analysis units, one specific MS ² spectrum, all MS ² spectra obtained from one test sample, all MS ² spectra obtained from all test samples measured in one experiment, etc. Any one of the above may be selected. Further, a plurality of MS ² spectra particular in the MS ² spectra obtained from one of the test sample may be designated.
All of the MS ² spectra specified as one analysis unit in step S1 are subjected to subsequent analysis.

  When the analysis is started, the database search unit 21 reads out the spectrum data included in the analysis unit designated in step S1 from the storage unit 20 (step S2), and performs a normal database search using the protein database 22. Thus, the peptide is estimated and the protein is identified (step S3). Then, a list of identification results, that is, information on the identified peptides and proteins is displayed on the screen of the display unit 4 via the display processing unit 27 (step S4). The analyst confirms this list, determines, for example, whether or not the attribution is reliable based on a score value indicating the reliability of identification, and inputs the determination result from the input unit 3 (step S5).

  When it is input that the identification result is reliable, the search result extraction unit 23 stores the identification result in the confirmation result storage unit 25 (step S6). On the other hand, if it is input that the identification result is not reliable, the search result extraction unit 23 stores the identification result in the non-confirmed result storage unit 24 (step S7). That is, by the processing of steps S5 to S7, whether the identification result obtained by the database search is deterministic that is reliable or non-deterministic that is not reliable is discriminated based on the judgment of the analyst. Stored in the storage unit. Then, it is determined whether or not the processing in steps S2 to S7 has been performed for all the mass spectra included in the analysis unit specified in step S1 (step S8). If there is an unprocessed mass spectrum, the processing proceeds from step S8 to S2. And the process of steps S2 to S8 is repeated.

Therefore, when all the mass spectra for a plurality of samples are designated as one analysis unit in step S1, a database search is executed based on all the mass spectra obtained for each sample, and the identification result That is, information indicating the estimated peptide or protein is stored in one of the two storage units 24 and 25.
Since the identification result stored in the determination result storage unit 25 by the above process is highly reliable, it is not necessary to perform a manual attribution determination process with the judgment of the analyst. On the other hand, since the identification result stored in the non-confirmed result storage unit 24 has low reliability, a manual attribution determination process with an analyst's judgment is performed.

  That is, the search result integration unit 26 extracts all protein candidates stored as non-deterministic results for one analysis unit, and information about the peptides estimated based on different mass spectra for each protein entry. Aggregate. Then, via the display processing unit 27, the amino acid sequence corresponding to the estimated peptide in the amino acid sequence of the protein is displayed on the display unit 4 in a state that is clearly identified in a form that can be visually distinguished from the other parts. (Step S9). That is, when one peptide estimated to exist based on a certain mass spectrum has an amino acid sequence contained in proteins a, b, and c, for each of the proteins a, b, and c The position of the corresponding peptide in the amino acid sequence of the protein is displayed in an easy-to-understand manner. At the same time, other peptides identified from the same sample, which may be included in the same protein a, b, c, are also clearly indicated in the same amino acid sequence. That is, all peptides obtained from one analysis unit that can be assigned to the protein are specified in the amino acid sequence. Typically, in the notation of the amino acid sequence of a protein (so-called three-letter notation or one-letter notation), the amino acid sequence corresponding to the estimated peptide may be displayed in a different character color.

  In addition, when estimating a peptide in the process of database search, a score indicating the reliability of the estimation is given, and if the score value is extremely high, it can be considered that the peptide surely exists. . Therefore, when distinguishing between these deterministic peptides and non-deterministic peptides that are not (or score values are low or not so high) and clearly indicate the estimated peptide in the amino acid sequence of the protein, It is preferable to display such that the peptide can be identified as deterministic or non-deterministic.

Here, a specific display example will be described with reference to FIG.
Now, it is assumed that four types of mass spectrum (MS ² spectrum) data A, B, C, and D are included in a data group designated as one analysis unit. This is the case, for example, when four different mass spectra A, B, C, and D are obtained as a result of measuring one sample under the same conditions. By database search based on mass spectrum data, mass spectrum B is definitely assigned to a peptide derived from a specific protein b, and mass spectrum D is assigned to a peptide derived from a specific protein c, while mass spectrum A and mass spectrum are assigned. Suppose that no definitive attribution result was obtained from C. In the database search based on the mass spectrum A, it is assumed that three proteins a, b, and c are found as candidates with low scores, that is, as uncertain results. In this case, if the mass spectrum A is manually identified, for example, a display as shown in FIG. 3 is made.

  3 (a) to 3 (c) show simplified amino acid sequences for the sake of convenience (actual display is 3-letter code or single-character code of amino acid sequence), and dotted lines are recorded in the protein database 22. The amino acid sequence of the protein is shown as a solid line, and the amino acid sequences of the peptide candidates that are indeterminately assigned from the mass spectrum other than A belonging to the same analysis unit as the mass spectrum A are shown. A wavy line indicates an uncertain or non-deterministic peptide candidate estimated by a database search based on a mass spectrum to be manually identified (ie, mass spectrum A in this case), and a double wavy line indicates the same data. Peptides determined to be deterministic by database search based on another mass spectrum in the group (ie, in this case, the peptide estimated from mass spectrum B for protein b and the mass spectrum D for protein c) Peptide). Therefore, the solid line is an undefined peptide candidate derived from the mass spectrum C. That is, this display format itself can be regarded as the same display format as the above-described existing Protein view in Mascot, and the wavy line corresponds to a portion displayed in red in the Mascot Protein View. The symbol with P in the inverted triangle indicates post-translational modification phosphorylation, and the symbol with G in the inverted triangle also indicates the post-translational modification sugar chain.

As shown in FIG. 3A, four peptides are assigned to protein a. According to the notation definition described above, these four peptides are derived from mass spectra other than mass spectrum A (ie B, C, D), and these mass spectra are MS ² spectra, so that It can be seen that it is derived from spectrum C. One of the four peptides is presumed to have a post-translational modification of phosphorylation, but these four are definitive, ie not included in proteins other than the protein This does not include any peptides that have been confirmed from the genome information of the organism.

  On the other hand, as shown in FIGS. 3B and 3C, protein b and protein c each contain a definitive peptide (derived from mass spectrum B or mass spectrum D). It can be inferred that it was included in the samples measured in the data group. Therefore, it can be said that there is a higher possibility that the assignment to protein b and protein c is the correct analysis than the assignment to protein a. Of course, there is no sufficient basis to conclude that there is no data derived from protein a in the data group designated as the analysis unit. However, as long as the presence of the protein b and the protein c is clear, it is desirable that the peptide whose presence is clarified by the measurement is preferentially assigned to the protein b and the protein c over the protein a.

  It is necessary to use other information to determine whether it is appropriate to assign the mass spectrum to be manually identified to protein b or protein c. For example, there are many hits in the amino acid sequence of the protein, that is, the assignment to the final protein is determined according to a criterion such as judging that the peptide with a high assignment to the whole protein is the correct answer. The decision may be made. This is a method similar to pathway analysis. That is, the correct answer is determined by whether or not there is a series of related hits rather than a single hit. Conventionally, the cover rate has been used to make such a determination. However, according to the system of this example, the state (distribution) in which the estimated peptide covers the protein is visually confirmed. Thus, it can also be determined whether or not it is appropriate to use the coverage ratio for the determination. Therefore, the accuracy of judgment is improved.

  Further, if only two types of protein candidates, protein a and protein b, shown in FIG. 3 are obtained by visual display supporting such manual identification work, the mass spectrum that is the object of manual identification is obtained. Can be easily determined to be attributed to the protein b whose existence is already definite.

  In the system of this example, the display of the amino acid sequence for each protein candidate as described above is displayed for all protein candidates for which indeterminate results are obtained by database search, that is, the ranking of search results and the presence or absence of post-translational modification Do it regardless of any. Then, the analyst confirms the display on the screen of the display unit 4, selects a result that seems to be optimal, and gives an instruction from the input unit 3 (step S10). In response to this instruction, the result determination processing unit 28 determines the final attribution result including the definite protein attribution result, and the final analysis result is displayed on the screen of the display unit 4 via the display processing unit 27. (Step S11). In other words, the final analysis results are automatically assigned by the database search, and the results of the database search are indeterminate, so the judgment by the analyst is added and the attribution is manually determined. It will be mixed.

[Modification of the first embodiment]
In the protein identification system of the first embodiment, in addition to the display as described above, a display is provided so that an analyst can manually distinguish between a peptide whose assignment has been manually determined and a peptide whose assignment has been automatically determined. You may make it do.
For example, the assigned peptide candidates shown in solid lines in FIG. 3 are displayed in blue, and the definitive peptides are displayed in black bold (bold), corresponding to the current mass spectrum that is the object of manual identification. When the peptide to be displayed is displayed in red, the peptide to which the assignment has been determined by the manual identification work performed in the analysis unit so far may be clearly indicated with an underline. This allows the analyst to visually grasp which amino acid sequence range of the displayed protein candidates has been assigned deterministically or manually. The efficiency of identification work is expected to be improved. In addition, it is easy to review the result of manual identification already performed by looking over the entire amino acid sequence. Since the result of manual identification already performed can be confirmed in this way, an identification result closer to the correct answer can be obtained on the basis of a human (analyst) judgment criterion.

[Second Embodiment]
Next, a protein identification system according to a second embodiment of the present invention will be described with reference to the accompanying drawings.
FIG. 4 is an overall configuration diagram of the protein identification system of the second embodiment, and FIG. 5 is a flowchart showing an identification processing procedure in the protein identification system of the second embodiment. In FIG. 4, the same components as those of the system of the first embodiment shown in FIG.

  In the protein identification system of the second embodiment, the data analysis unit 2 includes a database search unit 21, a protein database 22, a search result extraction unit 203, a related information search / calculation unit 204, a related information database 205, a result integration unit 206, and a display. A processing unit 207 and a result confirmation processing unit 208 are included as functional blocks.

  In the related information database 205, various types of data that can be used for protein identification by methods other than protein identification using mass spectrum data are recorded in association with the protein as related information. Specifically, as related information, for example, analysis results of gene expression information obtained by performing microarray analysis of genes corresponding to various proteins (that is, genomics analysis information), known pathway maps (genomics analysis) , Such as pathway maps indicating gene expression networks and pathway maps indicating metabolite networks in metabolomics analysis), and three-dimensional structure information of various proteins obtained by X-ray structural analysis. Note that the related information database 205 is not necessarily a database prepared by the user, but may be obtained on-demand necessary information from a plurality of external databases by using Semantic Web technology or the like.

An example of a protein identification procedure using the protein identification system of the second embodiment will be described with reference to FIG.
It is assumed that the spectrum data storage unit 20 stores mass spectrum data obtained by performing MS ² analysis in the mass analysis unit 1 on a test sample including a peptide mixture derived from the target protein.

  When the analysis is started according to the analyst's instruction, the database search unit 21 reads out the spectrum data designated by the analyst from the storage unit 20 (step S21), and performs a normal database search using the protein database 22. Thus, the peptide is estimated and protein identification is attempted (step S22). Then, a list of identification results, that is, information on the identified peptides and proteins is displayed on the screen of the display unit 4 via the display processing unit 207 (step S23). Steps S21 to S23 are the same processes as steps S2 to S4 of the first embodiment.

  This list of identification results includes the assigned peptide candidates and protein candidates together with the scores given in the database search process. The analyst confirms this list, and, for example, the identification result determined that the score value indicating the reliability of identification is low and the reliability of the attribution is poor is input to the input unit 3 as an object of identification processing characteristic of the present embodiment. (Step S24). When a plurality of uncertain protein candidates are obtained by database search based on a certain mass spectrum, all of these candidates may be selected. In addition, a plurality of identification results respectively obtained based on a plurality of mass spectra may be collected and selected as a target of identification processing characteristic of the present embodiment.

  The search result extraction unit 203 extracts the identification result selected in step S24 and creates a list (step S25). The related information search / calculation unit 204 accesses the related information database 205 for each of the identification results shown in the list created in step S25, and searches for omics information and other related information that can be used to determine attribution. (Step S26).

When the relevant information is found in the relevant information database 205, the relevant information search / calculation unit 204 attempts to attribute the uncertain protein candidate using the relevant information. In other words, an attribution determination process in which identification by database search based on mass spectrum data and related information such as omics information are integrated is performed (step S27). Since the method of the attribution determination process varies depending on the type of related information, it will be described in detail later. As described above, since the related information is information that is not directly related to the mass spectrum data, the attribution determination process narrows down the uncertain identification result obtained by the database search from another viewpoint, and the accuracy of attribution. Can be raised.
Here, a specific example of related information and attribution analysis processing using the related information will be described.

[When using microarray analysis results as related information]
One of the relevant information useful in determining protein assignment is information on the expression level of genes obtained by microarray analysis. In recent years, when a biological tissue or cells are measured using a mass spectrometer, microarray analysis is often performed on the same biological tissue or equivalent cells. Therefore, the microarray analysis result may be stored in advance in the related information database 205 and used when determining protein attribution based on the mass spectrum. That is, as a method of the attribution determination process, a protein corresponding to mRNA having a large detection amount from the microarray analysis result may be attributed preferentially. In the microarray analysis for such purpose, it is only necessary to measure the relative amount of mRNA. Therefore, the analysis result by the RNA sequencing method using a so-called next-generation sequencer (NGS) may be used as related information.

[When using pathway maps as related information]
For example, when a pathway map showing a gene expression network or a pathway map showing a metabolite network in metabolomics analysis is available as related information, all candidate proteins with low reliability listed in the identification result list are all , Each of them is mapped onto various known pathway maps. At this time, since the pathway map in which the most protein candidates are mapped is considered to have the highest reliability, a protein corresponding to the pathway map may be selected as an assignment result.

  For proteins assigned with a sufficiently high score by database search, the protein may be mapped in all pathway maps including the corresponding protein prior to analysis using the pathway map. In addition, all uncertain protein candidates such as scores that are close to each other are mapped onto the pathway, and the pathway map with the most mapped proteins and the most connected routes is adopted. The protein mapped to can be assigned to the designated mass spectrum.

  In normal proteomic analysis, a difference analysis or an enrichment analysis may be performed to extract a list of characteristic proteins. Such information may be used as related information. In particular, when identifying a biological substance whose pathway map is not sufficiently prepared, enrichment analysis is useful instead of mapping on the pathway map.

  Of course, the microarray analysis result and the pathway map may be used together as related information. That is, a microarray analysis result (or NGS analysis result) corresponding to a mass analysis result is mapped on a pathway map, and a pathway having a high possibility of expression is estimated. If the protein expression obtained based on the mass spectrum is mapped there, the expression level of mRNA and the abundance of the protein are not proportional (for example, when the protein is transported to another place in vivo) May be able to select an appropriate protein candidate.

FIG. 6 is a schematic diagram showing a specific analysis example in this case. FIGS. 6A to 6C are examples of pathway maps, where each node indicated by a circular symbol ◯ indicates a protein or its gene, and an edge indicated by an arrow indicates the progress direction of the reaction.
Here, from the pathway maps showing a number of known gene expression networks, both the protein attribution results (indeterminate and uncertain) and the corresponding microarray analysis results for all mass spectra to be analyzed A pathway map is selected, and a node indicating a protein assigned by database search and a node corresponding to a gene detected to have a transcriptional change by microarray analysis are marked. In this example, halo is added to the node corresponding to the assigned protein, and a seven-pointed star is added to the node corresponding to the detected gene.

  In FIG. 6 (a), there is no overlap between the seven-pointed star indicating mRNA expression and the halo symbol indicating protein expression. Such a state is almost impossible in an actual cell, and it is very unlikely that the phenomenon shown in this pathway map has occurred. In FIG. 6 (b), the symbol of the seven-pointed star and the symbol of halo overlap at only one node indicated by A, but the number of halos is small as a whole, and this protein was actually detected by this measurement. It is still unclear. In FIG. 6C, there is a pathway in which the symbol of the seven-pointed star and the symbol of halo overlap at four nodes, and are generally connected from the upper left start point to the lower right end point. Therefore, from the measurement data of both mRNA and protein, it can be inferred that the path from the upper left starting point to the lower right end point in the pathway map is actually expressed. In such a case, a protein with a halo symbol on the pathway may be assigned to the corresponding mass spectrum.

The use of the pathway map as described above is particularly useful in an imaging mass microscope.
For example, it is already known that specific lipids increase in cancer cells, and if the expression level of all genes on the pathway map of lipid synthesis system or fatty acid synthesis system is increased, Increased reliability of corresponding protein assignment results. In general, in a mass spectrometer using a MALDI ion source, measurement is performed in a state where substances are mixed even more than in a liquid chromatograph mass spectrometer in which a liquid chromatograph is coupled to the previous stage of the mass spectrometer. That is, since the results of one experimental lot are measured together, it can be estimated that the above-described methods of pathway analysis and enrichment analysis are effective.

[When using protein structure information as related information]
The microarray analysis results and pathway maps described above are omics information obtained by genomics analysis and metabolomics analysis other than proteomics analysis, but protein three-dimensional structure information and the like can also be used as related information as information other than omics information.

  FIGS. 8A and 8B are explanatory diagrams of protein attribution determination processing using the three-dimensional structure information of the protein as related information. As in FIG. 3, the amino acid sequences of protein candidates are simply shown. It is a thing. Similar to FIG. 3, the dotted line is the amino acid sequence recorded in the database, and the solid line is the peptide candidate identified by the database search based on the mass analysis result. In addition, the shaded area along the amino acid sequence is the interior of the protein based on the known protein structure information (more precisely, it is exposed to the external water environment). No). In the example of FIG. 8 (a), when the assigned peptide and the three-dimensional structure information of the protein are combined, at the site where the inside of the protein, possibly a hydrophobic core, is formed and at least blocked from the outside water environment. Phosphorylation has occurred, but this is difficult to think from the technical common sense. As a result, it can be determined that the possibility that the attribution as shown in FIG. 8A is correct is very low.

  On the other hand, in the case of the example shown in FIG. 8B, there is a possibility that the peptide subjected to the sugar chain modification does not exist inside the protein and is exposed to the external water environment. This is a result consistent with the presence of sugar chains. Therefore, when FIG. 8A and FIG. 8B are compared, it can be determined that the probability that the example of FIG. 8B can occur is higher.

  The three-dimensional structure information of the protein as described above can be acquired from the protein structure data bank (PDB) when the structure is known. On the other hand, if the structure is unknown, the three-dimensional structure of the protein candidate may be estimated by an appropriate technique such as the 3D-1D method and mapped onto the amino acid sequence.

  Returning to FIG. As described above, by using omics information known at the time of protein assignment work and other related information, mass spectra that were difficult to assign by database search etc. are known with a certain basis. The possibility of assigning to a protein increases. Therefore, if the protein candidates are narrowed down to one or more, the result is displayed on the screen of the display unit 4. In order for the analyst to finally select an appropriate result using the related information, it is preferable to display a result obtained by integrating the attribution result by the database search and the related information. Therefore, the result may be displayed in the format shown in FIG. 6 or the format shown in FIG. 8 according to the type of related information used. The analyst confirms this result on the screen, and if there are still a plurality of candidates, selects the candidate that seems to be optimal (step S28). The result confirmation processing unit 208 confirms the final identification result according to the selection of the analyst, and displays on the screen of the display unit 4 together with the protein or peptide to which it belongs, as well as related information used for the attribution determination processing. (Step S29).

  In addition, although the said 2nd Example gave as an example the case where a protein is identified as a biological substance, the same attribution determination method is applied also to compounds, lipids, carbohydrates, etc. other than the protein derived from a biological body. be able to.

  For example, since there is a database such as MassBank for sugar chains, lipids, or other biological compounds, for example, by database search based on mass spectra obtained for samples by using such a database, Candidate candidates can be identified with the same degree of reliability as proteins. If the compound candidates obtained in this way are mapped to a pathway map of metabolites obtained by metabolomics analysis or the like, metabolites can be selected just like proteins.

  When it is desired to identify such a non-protein compound, a pathway that is likely to be expressed may be estimated by referring to the pathway that produces the compound and using the amount of mRNA of the enzyme involved.

FIG. 7 shows an example of a pathway map showing the metabolism of a compound in the same notation as FIG. In a pathway map showing the metabolism of a compound, a compound is generally assigned to a node, and an enzyme that manages a chemical reaction necessary to generate the compound, that is, a protein is generally represented by an arrow edge. Therefore, the same notation is used in this pathway map. That is, when a certain compound exists as a result of the database search, the symbol of a seven-pointed star is superimposed on the node, and when a protein exists, a halo symbol is displayed on the edge.
In the example of FIG. 7, the lightly shaded pathway is generally connected by mRNA and protein from the upper left start point to the lower right end point of the pathway map shown in FIG. Therefore, it is considered that this pathway is highly likely to be manifested. Therefore, it can be determined that it is appropriate to assign a protein displaying halo corresponding to the route.

  The above-described embodiments are merely examples of the present invention, and it is a matter of course that modifications, corrections, additions, and the like are appropriately included within the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 ... Mass analysis part 2 ... Data analysis part 20 ... Spectral data storage part 21 ... Database search part 22 ... Protein database 23, 203 ... Search result extraction part 24 ... Unconfirmed result storage part 25 ... Confirmation result storage part 26 ... Search result Integration unit 27, 207 ... Display processing unit 28, 208 ... Result determination processing unit 204 ... Related information search / calculation unit 205 ... Related information database 3 ... Input unit 4 ... Display unit

Claims (14)

An identification support method for a biological substance that supports identification work for identifying a target protein based on the result of mass spectrometry of a sample containing a peptide mixture obtained by fragmenting a target protein that is a biological substance,
a) Candidates that estimate the amino acid sequences of at least some of the peptides contained in the sample based on mass spectral data obtained by mass spectrometry of the sample and extract a plurality of protein candidates as target proteins based on the estimation result An extraction step;
b) For each of a plurality of protein candidates extracted in the candidate extraction step, the amino acid sequence of the peptide estimated in the candidate extraction step can be distinguished from other amino acid sequences in the description of the amino acid sequence of the protein. A display processing step to display as it is,
A method for assisting identification of a biological substance, characterized in that, by referring to the display result in the display processing step, the analyst can select a candidate from among protein candidates. A method for supporting identification of a biological substance according to claim 1,
An analysis unit specifying step in which an analyst specifies a range of data to be reflected in one analysis as one analysis unit;
In the candidate extraction step, the estimation of the amino acid sequence of the peptide and the extraction of protein candidates are performed based on all data designated as one analysis unit,
In the display processing step, the estimation result of peptides based on all data designated as one analysis unit is integrated and displayed for each protein candidate, and the identification support method for a biological substance is characterized. A biological substance identification support device that supports identification work for identifying a target protein based on the result of mass spectrometry of a sample containing a peptide mixture obtained by fragmenting a target protein that is a biological substance,
a) Candidates that estimate the amino acid sequences of at least some of the peptides contained in the sample based on mass spectral data obtained by mass spectrometry of the sample and extract a plurality of protein candidates as target proteins based on the estimation result An extractor;
b) For each of a plurality of protein candidates extracted by the candidate extraction unit, the amino acid sequence of the peptide estimated by the candidate extraction unit can be distinguished from other amino acid sequences in the description of the amino acid sequence of the protein. A display processing unit for displaying as it is,
An identification support apparatus for biologically derived substances, characterized in that an analyst can select a candidate from protein candidates by referring to a display result by the display processing unit. An identification support apparatus for a biological substance according to claim 3,
An analysis unit designating unit for the analyst to specify the range of data to be reflected in one analysis as one analysis unit,
The candidate extraction unit performs estimation of peptide amino acid sequences and extraction of protein candidates based on all data designated as one analysis unit,
The display processing unit integrates and displays the estimation results of peptides based on all data designated as one analysis unit for each protein candidate, and the identification support apparatus for a biological substance. A method for assisting identification of a biological substance that supports identification work for identifying the target biological substance based on the result of mass spectrometry of a sample containing the target biological substance,
a) a candidate extraction step of extracting a plurality of candidate substances of a target biological substance based on mass spectral data obtained by mass spectrometry on a sample;
b) a substance selection step of selecting a highly probable substance candidate from among the plurality of candidate substances using reference information related to the target biological substance that cannot be derived from the mass spectrum data;
A method for assisting identification of a biological substance, characterized by comprising: A method for assisting identification of a biological substance according to claim 5,
The target biological substance is a protein, and the reference information is omics information other than the proteomic analysis result by mass spectrometry. It is the identification assistance method of the biological substance of Claim 6, Comprising:
The reference information is information on gene expression level by genomics analysis, and a method for assisting identification of a biological substance, characterized in that: A method for assisting identification of a biological substance according to claim 5,
The reference information is a pathway map showing a gene expression network in genomics analysis, or a pathway map showing a metabolite network in metabolomics analysis. A method for assisting identification of a biological substance according to claim 5,
The biological support substance identification support method, wherein the target biological substance is a protein, and the reference information is three-dimensional structure information of the protein. A biological material identification support device that supports identification work for identifying a target biological material based on the result of mass spectrometry of a sample containing the target biological material,
a) a candidate extraction unit that extracts a plurality of candidate substances of a target biological substance based on mass spectral data obtained by mass spectrometry of a sample;
b) using a reference information related to the target biological substance that cannot be derived from the mass spectrum data, a substance selection unit that selects a highly probable substance candidate from the plurality of candidate substances;
A biological material-derived substance identification support apparatus comprising: It is an identification assistance device for a biological substance according to claim 10,
The target biological substance is a protein, and the reference information is omics information other than the proteomic analysis result by mass spectrometry. It is an identification assistance apparatus of the biological substance of Claim 11, Comprising:
The reference information is information on a gene expression level by genomics analysis, an identification support apparatus for a biological substance, characterized in that: It is an identification assistance device for a biological substance according to claim 10,
The reference information is a pathway map showing a gene expression network in genomics analysis, or a pathway map showing a metabolite network in metabolomics analysis. It is an identification assistance device for a biological substance according to claim 10,
An identification support apparatus for a biological substance, wherein the target biological substance is a protein, and the reference information is three-dimensional structure information of the protein.

Images