JP4855780B2 - Peptide identification method and identification apparatus using mass spectrometry - Google Patents

Description

  The present invention provides a method for mass-analyzing a sample containing a peptide mixture and identifying the amino acid sequence of each peptide using mass spectral data obtained thereby, and for performing such identification mainly using a computer. Relates to the device.

In recent years, protein structures and functions have been rapidly analyzed as post-genomic research. As one of such protein structure / function analysis methods (proteome analysis), protein expression analysis and primary structure analysis using mass spectrometers are widely performed, and a quadrupole ion trap So-called MS ⁿ analysis (n is an integer of 2 or more), which captures and cleaves a specific peak by, for example, collision induced decomposition (CID), is effective. In general, in MS ² (= MS / MS) analysis, first, an ion having a specific mass ( strictly m / z) is selected from an analysis object as a precursor ion, and the precursor ion is cleaved by CID. Then, information on the mass and chemical structure of the target ion can be obtained by mass analysis of ions (product ions) generated by cleavage.

When the amino acid sequence of a protein is determined by MS ⁿ analysis as described above, the protein is first digested with an appropriate enzyme to obtain a mixture of peptide fragments, and then the peptide mixture is subjected to mass spectrometry. At this time, since stable isotopes having different masses exist in the elements constituting each peptide, a plurality of peaks having different masses are generated even if the peptides have the same amino acid sequence due to the difference in the isotopic composition. The plurality of peaks are composed of an ion (main ion) peak composed only of the isotope having the maximum natural abundance ratio and an ion (isotope ion) peak containing other isotopes, which are spaced by 1 Da. A peak group consisting of a plurality of peaks arranged in a row is formed.

Subsequently, from a mass spectrum data of the peptide mixture as described above, a set of isotope peaks derived from a single peptide is selected as a precursor ion, and ions obtained by cleaving the precursor ion ( Product ion) mass analysis (MS ² analysis).

  The amino acid sequence of the test peptide can be determined by subjecting the spectral pattern of the product ion obtained as described above and the spectral pattern of the precursor ion to a database search. Standard software for identifying peptides based on mass spectral data obtained by mass spectrometry is conventionally used. For example, as a typical database search engine for proteome analysis, MASCOT provided by Matrix Science is well known.

A schematic analysis procedure when this software is used is shown in FIG. That is, as a key mass has appeared in the mass spectrum obtained fragment ions, searches the peptide mass matches among the peptides that are registered in the database (step S1), the search of a plurality of obtained Candidate peptide names are ranked and listed in order of match (step S2).

Analysts estimate peptides by referring to this list, but in fact, the highest ranking in the list of candidate peptides is often not the correct substance, and the above software search is not necessarily reliable. Is not expensive. The reason for this is considered to be that although matching of fragment ion mass is considered in the above software, matching with the peak intensity of fragment ions is not considered at all or hardly.

As an example of the analysis result using MASCOT, FIG. 17 shows an example in which a peptide that is not a correct substance is ranked first on the list. This example is based on the mass spectrum obtained as shown in FIG. 16 by mass spectrometry of angiotensin II (an aminotensin II: amino acid sequence = [DRVYIHPF], mass = 1,046.5 Da). This is a list of candidates. As a result, the correct amino acid sequence has been ranked in the fourth rank among those listed as candidates. In this example, the degree of matching is expressed as a score, but the peptide with the amino acid sequence [DFHLDEDR] obtained the highest score (31 points), and the correct answer Angiotensin II scored 28 points. Stay in 4th place. However, the scores are close from 1st to 4th, and it is difficult to conclude which of these is correct from this result.

Conventionally, a database search engine called SEQUEST provided by Thermo Electron is used separately from MASCOT. This software takes into account the matching of both the fragment ion mass and its peak intensity. However, basically, just like MASCOT, the database has only mass- related data, and the strength is very simple (specifically, the b / y ion intensity is 50, Dehydration / desalting ionic strength is only evaluated by matching with mass spectral data obtained by measurement by giving fragment ionic strength of candidate peptides under 10). Therefore, the reliability cannot be expected with respect to the intensity matching.

  FIG. 18 shows an example of evaluation using SEQUEST based on the mass spectrum data shown in FIG. In this result, the correct answer Angiotensin II [DRVYIHPF] is ranked first, and [DFHLDEDR], which was mistakenly ranked first in MASCOT, is second. However, both scores are very close at 0.3067 and 0.2977, and it is difficult to conclude which is the correct answer among the peptide molecules ranked from 1st to 3rd.

In general, when it is difficult to narrow down to one peptide by the identification process as described above, for example, the information of the fragment ion mass given as information is adjusted by reducing or increasing, and the identification is attempted again. Work is often done. However, such work is trial and error, and a reliable result is not always obtained, and there is a problem that throughput is nevertheless lowered.

On the other hand, recently, assuming a mathematical calculation model consisting of several tens of parameters, and providing a means to predict the spectrum pattern of fragment ion intensity under this assumption, the mass and intensity of fragment ions were combined. A method for performing matching has also been proposed (see Non-Patent Document 1). However, although the parameters used in this proposed method include physical implications such as proton affinity (an index of the force that attracts protons) and activation energy, they are limited by parameter fitting. It is merely a numerical parameter determined so as to fit as much as possible to a certain number of actually measured data. For this reason, this model formula is unreliable to apply to all types of peptides, and it is scientifically meaningful but has little practicality.

  Further, in Patent Document 1, when a peptide to be calculated is given in proteome analysis, the dissociation energy of each chemical bond of the amino acid sequence of the peptide is calculated, and the intensity corresponding to each dissociation pattern is calculated using the dissociation energy. A method has been proposed in which matching accuracy is improved by predicting and matching the predicted intensity value with mass spectrum data acquired by measurement. However, in this document, although the intensity (Intensity) corresponding to each dissociation pattern is calculated by calculating the reaction rate using dissociation energy, the specific method is not clarified.

  Actually, as will be described later, the reaction rate cannot theoretically be calculated using the dissociation energy, and the activation energy required to cleave at the molecular position, not the dissociation energy, to obtain the reaction rate. Must be used and this is not easily determined. Furthermore, other parameters such as the temperature and frequency factor of the molecule are necessary to calculate the reaction rate, and these are difficult to determine. For these reasons, the proposed method cannot be put to practical use. Even if it is possible to estimate the intensity from the dissociation energy, it is difficult to increase the estimation accuracy because an uncertain factor is added in the process, and when comparing this estimated intensity with the measured intensity, There may be a problem in reliability.

JP 2004-12355 A Z.Zhang, “Predication of Low-Energy Collision-Induced Dissociation Spectra of Peptides” Analytical Chemistry, Vol.76, No.14, July 15, 2004, pp.3908-3922

  The present invention has been made in order to solve the above-mentioned problems, and an object of the present invention is to provide a peptide identification method and identification apparatus capable of increasing the accuracy of determination of amino acid sequences of proteins and peptides based on mass spectral data. Is to provide.

1st invention made in order to solve the said subject is a peptide identification method which determines the amino acid sequence of the peptide mixture in a test sample based on the mass spectral data acquired by ^MSn analysis,
a) a primary candidate selection step of selecting a plurality of candidate peptides based on the mass spectrum data;
b) For each of the plurality of candidate peptides, a dissociation energy acquisition step for obtaining dissociation energy corresponding to all fragment ions;
c) Correlation for calculating a correlation coefficient between the dissociation energy obtained in the dissociation energy acquisition step and the actually measured fragment ion intensity based on the mass spectral data estimated to correspond to each of the plurality of candidate peptides. Investigation steps,
d) a secondary candidate selection step of narrowing down candidate peptides based on the correlation coefficient obtained for each candidate peptide;
It is characterized by performing.

A second invention made to solve the above problems is an apparatus embodying the first invention, wherein the peptide mixture in the test sample is analyzed based on the mass spectrum data obtained by MS ⁿ analysis. A peptide identification apparatus for determining an amino acid sequence,
e) primary candidate selection means for selecting a plurality of candidate peptides based on the mass spectrum data;
f) for each of the plurality of candidate peptides, dissociation energy acquisition means for obtaining dissociation energy corresponding to all fragment ions;
g) Correlation for calculating a correlation coefficient between the dissociation energy obtained by the dissociation energy acquisition unit and the actually measured fragment ion intensity based on the mass spectral data estimated to correspond to each of the plurality of candidate peptides. Investigation means,
h) secondary candidate selection means for narrowing candidate peptides based on the correlation coefficient obtained for each candidate peptide by the correlation investigation means;
It is characterized by having.

  As will be described in detail later, in order to obtain the reaction rate of the cleavage reaction, it is necessary to provide parameters such as activation energy, molecular temperature, and frequency factor necessary for cleavage at the molecular position. Since it is practically impossible, the reaction rate cannot be obtained as it is. However, the inventor of the present application, based on theoretical considerations, shows that when cleavage occurs at various bonds (molecular positions) in a certain peptide, the dissociation energy and the intensity (ion content) of fragment ions obtained thereby. A conclusion was drawn that the relationship would have a high linear correlation.

  Therefore, first, a plurality of candidate peptides having a high probability are selected based on the mass spectrum data by a primary candidate selection step using existing proteome analysis software (for example, the above-mentioned MASCOT). Then, for each of the candidate peptides, the dissociation energy acquisition step determines the dissociation energy corresponding to the fragment ion generated by the cleavage of the peptide (energy necessary for generating the fragment ion). Although acquisition of dissociation energy can be performed by calculation, since this calculation generally takes time, it is not practical to perform calculation one by one during the identification process. Therefore, a database in which the dissociation energy is calculated for each combination of peptide amino acid sequences and the molecular positions to be cleaved is prepared in advance, and the corresponding dissociation energy can be obtained by referring to the database.

Then, in the correlation investigation step, for each of the plurality of candidate peptides, a predetermined allowable error with respect to the dissociation energy obtained as described above and the mass of the fragment ion that is estimated to correspond to it, that is, the fragment ion generated by its cleavage. and fragment ion intensities of actual measurement of mass to match those that have been extracted from the mass spectral data in the range, calculates an index value, for example, a correlation coefficient showing the correlation. In the secondary candidate selection step, based on the index value obtained for each candidate peptide, for example, the only candidate peptide determined to have the highest correlation is selected as the most likely peptide.

  However, even if a database prepared in advance is used to determine the dissociation energy, the number of amino acids contained in the peptide is large, the number of types of peptides with different amino acid sequences is enormous, and the dissociation energy is calculated in advance for all of them. It is not very practical to leave. Therefore, as one aspect of the present invention, a peptide in which three or more amino acids are bound is modeled as a dipeptide composed of two amino acids sandwiching the binding portion of interest, and the original peptide is determined by the dissociation energy of the dipeptide. It is better to substitute dissociation energy.

  Thereby, dissociation energy is calculated in advance for a dipeptide based on a combination of all common amino acids, and a database is prepared. In the dissociation energy acquisition step, the dissociation energy for an arbitrary fragment ion of a candidate peptide is determined from the database. The dissociation energy of the dipeptide can be called and used.

According to the peptide identification method according to the first invention and the peptide identification device according to the second invention, not only information relating to the mass of the peak appearing in the mass spectrum but also information relating to the intensity can be effectively utilized, Amino acid sequences of proteins and peptides can be determined with high accuracy.

  In particular, according to a simplified identification method using a dipeptide model, a dissociation energy database can be prepared with a reasonable calculation time, and this database can be used regardless of the peptide to be identified. Thus, the dissociation energy can be quickly obtained and identification can be performed quickly. Thereby, the cost for protein and peptide identification can be reduced, and the throughput of identification processing can be improved.

  First, the basic principle of the peptide identification method according to the present invention will be described.

ここでｋはフラグメントイオン強度、Ａは頻度因子と呼ばれる前指数因子、Ｅ_aはその分子位置で開裂するのに必要な活性化エネルギー、Ｒは気体定数、Ｔは分子の温度である。即ち、反応速度を求めるためには、活性化エネルギー、頻度因子、温度といった各パラメータが必要であり、このことから、前述したように特許文献１に記載の方法では反応速度が求まらず、それ故に強度を推定するのも困難であることが分かる。 A reaction in which a molecular ion is cleaved at a certain molecular position is called a so-called unimolecular chemical reaction, and is one of general chemical reactions. According to the standard theory of chemical reaction, the rate of the reaction is similar to the equation called Arrheniius equation, and is the following equation (1).

Here, k is a fragment ionic strength, A is a pre-exponential factor called a frequency factor, E _a is an activation energy necessary for cleavage at the molecular position, R is a gas constant, and T is a temperature of the molecule. That is, in order to obtain the reaction rate, each parameter such as activation energy, frequency factor, and temperature is required. From this, the reaction rate cannot be obtained by the method described in Patent Document 1 as described above. It can therefore be seen that it is difficult to estimate the intensity.

This reaction rate is an amount proportional to the amount of the reaction product, that is, the total amount of fragment ions generated by cleavage at the molecular position. FIG. 3 is a schematic diagram showing a state of energy change at the time of cleavage of molecular ions. As an example, if the molecular ion is now XY ⁺ , cleavage occurs at the junction between X and Y, and as a result, two types of generation of X ⁺ fragment ion and Y ⁺ fragment ion can be considered. The sum of the amount (intensity) of fragment ions is an amount proportional to the reaction rate k shown in the above equation (1).

Therefore, if the intensity is determined separately for each of X ⁺ fragment ions and Y ⁺ fragment ions, the dissociation energy (energy from the state before the reaction to decomposition into individual fragment ions) Ex shown in FIG. It is necessary to consider Ey. The division ratios Rx and Ry should be the following formulas (2) and (3) considering the thermodynamic equilibrium.

ここでＡ’は比例係数である。 Using these split ratios, the intensities of X ⁺ fragment ions and Y ⁺ fragment ions are obtained by multiplying equation (1) by equation (2) or by multiplying equation (1) by equation (3). It is proportional to Since the denominators of the equations (2) and (3) are common, the intensities Ix and Iy of the X ⁺ fragment ion and the Y ⁺ fragment ion are the dissociation energies Ex and Ey as shown in the following equation (4). Can be used.

Here, A ′ is a proportionality coefficient.

即ち、フラグメントイオン強度Ｉと解離エネルギーＥとの間に次の(6)式の直線関係が成り立つことになる。

この関係は特定の分子位置について導かれた関係式であるが、他の分子位置で開裂が生じた場合、必ずしも前指数因子Ａ’が同じとは限らない。しかしながら、或る一つの分子に関する開裂であって開裂の分子反応メカニズムが類似しているものであれば、前指数因子Ａはそれほど大きく変わらないと考えて構わない。こうしたことを考慮すれば、(6)式よりもさらに一般的な関係を示す(7)式が成り立つものと考えられる。

この(7)式は、例えば１つの分子の様々な開裂位置での解離エネルギーを横軸にとり、縦軸にその開裂により生成されるフラグメントイオンの強度をとって相関グラフを作成すれば、大きな直線相関を示すことを意味している。 As can be understood by taking the natural logarithm of both sides of the equation (4), it can be expressed by the following equation (5) without distinguishing X and Y.

That is, the following linear relationship (6) is established between the fragment ion intensity I and the dissociation energy E.

This relationship is a relational expression derived for a specific molecular position, but when cleavage occurs at other molecular positions, the pre-exponential factor A ′ is not necessarily the same. However, if the cleavage is related to a single molecule and the molecular reaction mechanism of the cleavage is similar, the pre-exponential factor A may be considered not to change so much. Considering these facts, it is considered that equation (7), which shows a more general relationship than equation (6), holds.

This equation (7) is a large straight line, for example, by taking the dissociation energy at various cleavage positions of one molecule on the horizontal axis and taking the intensity of fragment ions generated by the cleavage on the vertical axis to create a correlation graph. It means to show correlation.

In this way, even if the reaction rate of the cleavage reaction itself cannot be calculated as described above, it is possible to identify a molecule in consideration of the fragment ion intensity by using the correlation between the dissociation energy and the fragment ion intensity as described above. It can be seen that it is. This is naturally applicable to peptide molecules. Therefore, after performing mass matching using existing database search software for proteome analysis, such as MASCOT, and narrowing down the number of candidate peptides to some extent, the dissociation energy of the candidate peptide can be calculated for all fragment ions. It is possible to create a correlation table or correlation diagram between the dissociation energy and the corresponding measured fragment ion intensity, and the correct answer is that the relationship is a strong linear correlation as shown in Equation (7). It can be estimated that the possibility is high.

As an example of the index value for determining the correlation, when variable X = dissociation energy and variable Y = logarithm of fragment ion intensity, the correlation coefficient defined as follows (Pearson product moment phase) The relation number r can be used.
r = (covariance between variable X and variable Y) / (standard deviation of variable X) × (standard deviation of variable Y) (8)
The correlation coefficient r takes a value in the range from −1 to 1, and it is said that there is a negative correlation when the value is negative, and there is a positive correlation when the value is positive. When the value is close to 0, there is almost no correlation, and when the value is 1 or -1, all points (X, Y) are on one straight line. Therefore, when a correct answer is obtained, it is considered that the negative correlation is high as shown in equation (7).

As described above, in principle, it is necessary to calculate dissociation energies at various cleavage positions of a certain molecule. Since calculating the dissociation energy one by one when performing data analysis is very time consuming and may not be practical, it is preferable to calculate the dissociation energy of all bonds in advance and create a database, It is preferable that the corresponding dissociation energy is derived from the database. However, although there are at most 20 types of common amino acids, for example, considering only the combination of 5 amino acid sequences, considering two types of b ions and y ions, the calculation target of dissociation energy is 20 ⁵ × 2 = 64 × 10 ⁵ and a huge number, and it takes a long time to calculate all of this.

  That is, in order to narrow down candidate peptides by a method based on the above-described principle within the practical range, it is necessary to reduce the amount of calculation required to create a database of dissociation energy. Therefore, a method is used in which three or more amino acid sequences sandwiched between the N-terminus and the C-terminus are simplified to a dipeptide that is modeled only by the combination of two amino acids.

  In the case of peptide molecules, the dissociation energy can be obtained from a simplified model using a dipeptide model, and the dissociation energy in a peptide having three or more amino acid sequences can be substituted with this value. As described above, since there are about 20 types of common amino acids, even if the dissociation energy of dipeptides is calculated for all combinations, the number is about 20 × 20 × 2 = 800. A database such as a table of dissociation energy can be prepared easily. As a result, for any type of peptide, the dissociation energy can be obtained immediately from the database, and the amino acid sequence can be determined quickly.

A method for obtaining the dissociation energy in the case of dipeptide modeling will be described. FIG. 4 is a diagram for explaining the principle of cleavage of a general peptide. A general peptide molecule ion ionized by proton (H ⁺ ) causes cleavage at a bond portion mainly called a peptide group, and y ion The peaks of two groups of fragment ions called b ions are obtained on the mass spectrum. As shown in FIG. 4 (a), when the peptide group is cleaved at the boundary indicated by y2-b6, as shown in FIG. 4 (b), the y2 fragment ion and b6 are shown on the mass spectrum. Each peak of fragment ions is observed.

  As shown in FIG. 5, this peptide group is modeled as a dipeptide composed of amino acids on both sides of the cleavage position. Then, for the proton ionized product, the dissociation energy of the peptide group is calculated for each of b ion generation or y ion generation, and the value is used as the dissociation energy of b ions or y ions.

  The dissociation energy can be obtained by molecular dynamic calculation based on the quantum mechanical wave equation. Specifically, since the dissociation energy is the amount of energy change from the state before the fragmentation reaction to the state decomposed into fragment ions, the state energy before and after the reaction may be calculated and obtained from the difference.

  In order to obtain this energy difference, it is necessary to know what kind of molecular reaction actually causes fragmentation. From past studies, it is known that fragmented molecular reactions occur by two major mechanisms. One is a mobile proton model and the other is a reaction mechanism called a remote proton model. The dipeptide model is modeled as follows for each of these two fragmentation models and the dissociation energy is calculated.

[1] In the case of the mobile proton model In the mobile proton model, fragmentation (cleavage) starts when the proton moves to reach a certain peptide group. That is, as shown in FIG. 6, the proton is initially present at the lowest energy position (in many cases, the N-terminal), but when kinetic energy is acquired through gas collision such as CID, the proton leaves that position. To move (i). Then, when the proton reaches N of a certain peptide group, cleavage begins (ii). In the case of a general amino acid sequence, it can be calculated by this model.

  FIG. 7 is an explanatory diagram of the principle of calculation of dissociation energy by a dipeptide model, and is an example in the case of ionization at the N-terminus. As schematically shown in FIG. 7 (a), with respect to the peptide XYUZ (X, Y, U, and Z are each an amino acid residue represented by a symbol), the proton initially present at the N-terminus is from the N-terminus. It leaves and reaches the peptide group between Y-Z (intermediate state). Thereafter, the bond between Y and Z is broken, and a proton is cleaved into a b ion with a proton on the Y side and a y ion with a proton on the Z side. To be precise, the dissociation energy should be calculated according to the reaction shown in FIG. 7 (a), but this is approximated by dipeptide modeling as shown in FIG. 7 (b). That is, in FIG. 7B, the dissociation energy between the peptide bonds YZ is calculated in a state where both the initial state and the intermediate state (intermediate state) are replaced with dipeptides.

  FIG. 8 shows an example in which the schematic diagram of the dipeptide model shown in FIG. FIG. 8 shows an example of the simplest dipeptide, glycyl-glycine, but the molecular reaction formulas of other dipeptides can be considered in the same manner. In FIG. 8, the reaction route (i) is a route in which y ions are fragment ions, and at the same time, a neutral molecule called aziridinone, which is a triangular ring molecule, is generated. On the other hand, in the b ion generation route of the reaction route (ii), an amino acid (Gly) is simultaneously generated as a neutral molecule. Since the b1 ion generated initially by this reaction has an unstable structure, it immediately dissociates into a stable a1 ion, but there is no significant difference in energy. Therefore, the calculation of the dissociation energy for b ions may be performed in a reaction path that generates stable a1 ions.

[2] In the case of the remote proton model Unlike the mobile proton model described above, the remote proton model is a state in which the proton itself is deeply bound to a side chain molecule of a specific amino acid such as a basic amino acid of arginine or histidine. This is a special fragment reaction that occurs when protons are not easily separated from the side chain by collision with CID gas. Such a reaction is likely to occur at a position where a specific amino acid such as a basic amino acid or an acidic amino acid is present.

  FIG. 9 is an estimation diagram of a fragmented molecular reaction when arginine and aspartic acid are contained in the same peptide. This figure shows G.Tsaprailis and 7 others, “Influence of Secondary Structure on the Fragmentation of Protonated Peptides” ”, Journal of American Chemical Society (J. Am Chem. Soc.), 1999, 121, pp5142-5154. It shows the reaction formula for ions.

In the example of FIG. 9, the proton is bonded to the imino group (NH =) of the arginine side chain to become NH ₂ ⁺ =, and is deeply linked to the side chain in this manner. As already described, even if the CID gas is collided in this state, the protons are not easily separated from the side chain, so that fragment ions are not easily generated except for aspartic acid. However, aspartic acid has a carboxyl group (—COOH) in its side chain, and cleavage begins by the interaction between this carboxyl group and an adjacent peptide group. As a result, molecular ions having a special pentagonal cyclic structure (b ions in this figure) and neutral peptides are generated. In this fragmentation reaction, the proton ions in the side chain of arginine do not move and remain in the bound state.

  The calculation of the dissociation energy in the above case can be obtained by calculating along the fragmentation reaction of FIG. 9 assuming that the proton is first in the side chain of arginine. Considering a dipeptide consisting of two amino acids sandwiching, the dissociation energy may be calculated based on FIG.

  In addition, when the energy given by the collision of CID gas etc. becomes large, the proton captured by the side chain of basic amino acids, such as arginine, will be able to move, In that case, it will consider with a mobile proton model. The dissociation energy in this case may be calculated by making a dipeptide model along the schematic diagram shown in FIG. That is, as shown in FIG. 10 (a), between the state of the peptide in which the side chain of the first arginine (R) is protonized and the state fragmented into the last b ion or y ion, A hypothetical state (a state in which the N-terminus is protonated) and an intermediate state (a state in which protons reach the peptide group) are sandwiched. Then, the energy change from the first state to the virtual state is divided into the first step, and the virtual state to the last fragmented state is divided into the second step. As shown in FIG. Set the model.

  The second step dipeptide modeling can be performed as in FIG. For the first step, as shown in FIG. 10B, in the dipeptide containing arginine (R), the energy required for the proton to move from the side chain of arginine to the N-terminus is calculated. It is assumed that the sum of the dissociation energy obtained in the first step and the dissociation energy obtained in the second step is the dissociation energy obtained.

  By using the dipeptidation method as described above, the number of fragment ions for which dissociation energy is to be calculated can be reduced, thereby reducing the time required to create the dissociation energy database.

  An identification apparatus for carrying out the above-described peptide identification method according to the present invention will be specifically described. The substance of this peptide identification apparatus is a computer, and the process for determining the amino acid sequence of a peptide, which will be described later, can be executed by operating new proteome analysis software on the computer.

FIG. 1 is a flowchart showing a schematic processing procedure of peptide identification by this apparatus. Step S1 and S2 are conventional proteome analysis database search software, a process that may be implemented by, for example, MASCOT, as a key the mass of fragment ions that are appearing in the mass spectrum obtained by mass spectrometry (MS ^n), A peptide with a matching mass is searched from the peptides registered in the database, and a plurality of peptides obtained by the search are ranked and listed in order of matching degree.

  After that, among the listed peptides, select a plurality of candidate peptides with clearly higher scores than others, and as described above, intensity matching by correlation between dissociation energy and measured fragment ion intensity for each candidate peptide The evaluation is performed (step S3), and the most likely candidate peptide is selected by comparing the correlation evaluation results (step S4). The new function by these steps S3 and S4 is combined with the function by the conventional standard database search software, so that an identification result having higher reliability than the conventional one can be expected.

FIG. 2 is a detailed flowchart of the processing contents of steps S3 and S4. That is, first, mass spectral data obtained by mass spectrometry (MS ⁿ analysis) is input, and n (n ≧ 2) candidate peptides narrowed down by steps S1 and S2 are input (step S11). Then, 1 is set to the variable N (step S12), and the Nth, that is, the first candidate peptide is set as a calculation target (step S13).

  Next, with respect to the candidate peptide, dissociation energy is obtained for all fragment ions (step S14). At this time, a dissociation energy database prepared in advance by dipeptide modeling as described above is used. That is, as described above, two amino acid sequences sandwiching the target cleavage position are fragmented, and the combination of the two amino acids and the type of ion (b ion or y ion) are given to the database. The corresponding dissociation energy is derived. By using a database prepared in advance, the dissociation energy can be obtained quickly.

After obtaining the dissociation energies of all fragment ions in the candidate peptide, set a predetermined tolerance for the theoretical fragment ion mass corresponding to each cleavage, and the fragment ions that are assumed to correspond from the measured mass spectral data Find the peak of and get its intensity. Then, the correlation coefficient as described above is calculated as an index value for evaluating the correlation by associating the dissociation energy with the fragment ion intensity (step S15).

  When the correlation coefficient is obtained for the first candidate peptide in this way, it is determined whether or not the variable N is n (step S16). If N does not match n, 1 is added to N. Then, N is newly set (step S17), and the process returns to step S13. And the process of said step S13-S15 is performed about the next candidate peptide, and a correlation coefficient is calculated | required. In this way, correlation coefficients for all candidate peptides are obtained. When the correlation coefficients of all n candidate peptides are obtained, Yes is determined in step S16 and the process proceeds to step S18. The correlation coefficient between the dissociation energy and the fragment ion intensity for each candidate peptide is compared, and the most linear correlation is obtained. The peptide with the highest probability of being the most correct is selected. The identification (amino acid sequence determination) result is displayed on the display unit (step S19), and the process is terminated.

  Next, in order to verify the effect of the peptide identification method and apparatus according to the present invention, as a result of identification by MASCOT on the mass spectrum data of Angiotensin II (DRVYIHPF) shown in FIG. 16, as shown in FIG. An example of ranking up to the first place will be given to show that the method according to the present invention can be identified accurately.

  11 and 12 show the dissociation energy calculated by the dipeptide model in the method according to the present invention and the measured fragment for the peptide [DFHLDEDR] ranked first in the identification by MASCOT and the correct answer Angiotensin II (DRVYIHPF), respectively. The comparison with the ionic strength is shown by dividing it into groups of b series ions and y series ions. The horizontal axis represents the number m of bm ions or ym ions, and the number is shown on the upper side of the graph. The lower side represents the amino acid sequence of the peptide and indicates at which amino acid sequence the bm ion or ym ion causes cleavage. On the other hand, the vertical axis represents dissociation energy (absolute value) and fragment ion intensity (relative value).

  From the results of FIG. 11 and FIG. 12, the state of dissociation energy is compared with the state of fragment ion intensity, and the dissociation energy of some of the peptides (for example, y5, y6, y7 ions) in the MASCOT evaluation first position It can be seen that the change and the fragment ion intensity change are extremely different.

  FIG. 13 is a graph showing the correlation between the dissociation energy on the horizontal axis and the logarithmic value of fragment ion intensity on the vertical axis based on the results of FIGS. 11 and 12. As can be seen from this graph, a negative linear correlation is clearly seen in Angiotensin II (DRVYIHPF) in FIG. On the other hand, [DFHLDEDR] ranked first in MASCOT rank cannot clearly read the correlation as seen in FIG.

  FIG. 14 is a diagram showing the result of obtaining the correlation coefficient by the equation (8) from FIG. 13 separately for b series ions and y series ions. In Angiotensin II (DRVYIHPF), both the b series ion and the y series ion show a large negative correlation having a correlation coefficient of −0.5 to −0.7. In contrast, in DFHLDEDR ranked first in MASCOT rank, a large correlation is observed in the b series ions, but the correlation coefficient of the y series ions is close to 0 and it can be determined that there is almost no correlation. Thus, according to the method of the present invention, it can be concluded that there is a high possibility that Angiotensin II (DRVYIHPF), which is the correct answer, is clearly the correct answer, not [DFHLDEDR] ranked first in MASCOT rank.

  It should be noted that the above embodiment is merely an example of the present invention, and it will be understood that the present invention is encompassed in the scope of the claims of the present application even if appropriate modifications, corrections, additions, etc. are made within the scope of the present invention.

The flowchart which shows the rough process sequence of the peptide identification by the peptide identification apparatus of one Example of this invention. The detailed flowchart of the processing content of step S3 and S4 in FIG. The schematic diagram which shows the state of the energy change at the time of cleavage of a molecular ion. The schematic diagram for demonstrating the principle of cleavage of a general peptide. The figure which shows an example of dipeptide modeling. Explanatory drawing of the mechanism of cleavage in a mobile proton model. Explanatory diagram of calculation principle of dissociation energy by dipeptide model (in case of mobile proton model). The figure which shows an example which represented the schematic diagram of the dipeptide model with the specific molecular reaction formula. Estimated fragmentation reaction when arginine and aspartic acid are contained in the same peptide (in case of remote proton model). Explanatory drawing of calculation principle of dissociation energy by dipeptide model (in case of mobile proton model when basic amino acid is included). The figure which shows the result of having compared with the dissociation energy calculated by the dipeptide model in the method by this invention, and the measured fragment ion intensity about the peptide [DFHLDEDR] divided into the group of b series ion and y series ion. The figure which shows the result of having performed the same comparison as FIG. 11 about AngiotensinII (DRVYIHPF). FIG. 13 is a graph showing the correlation between the dissociation energy on the horizontal axis and the logarithmic value of fragment ion intensity on the vertical axis based on the results of FIGS. 11 and 12. The figure which shows the result of having calculated | required the correlation coefficient from FIG. 13 divided into b series ion and y series ion. The flowchart which shows the rough analysis procedure of the peptide identification using the existing software for proteome analysis. The figure which shows the mass spectrum obtained by mass-analyzing angiotensin II. The figure evaluated using MASCOT based on the mass spectrum data shown in FIG. The figure which shows the example evaluated using SEQUEST based on the mass-spectrum data shown in FIG.

Claims (5)

A peptide identification method for determining an amino acid sequence of a peptide mixture in a test sample based on mass spectral data acquired by MS ⁿ analysis,
a) a primary candidate selection step of selecting a plurality of candidate peptides based on the mass spectrum data;
b) For each of the plurality of candidate peptides, a dissociation energy acquisition step for obtaining dissociation energy corresponding to all fragment ions;
c) Correlation for calculating a correlation coefficient between the dissociation energy obtained in the dissociation energy acquisition step and the actually measured fragment ion intensity based on the mass spectral data estimated to correspond to each of the plurality of candidate peptides. Investigation steps,
d) a secondary candidate selection step of narrowing down candidate peptides based on the correlation coefficient obtained for each candidate peptide;
The peptide identification method using mass spectrometry characterized by performing these.   A peptide in which three or more amino acids are bonded is modeled as a dipeptide consisting of two amino acids sandwiching the binding portion of interest, and the dissociation energy of the dipeptide replaces the dissociation energy of the original peptide. A peptide identification method using mass spectrometry according to claim 1.   Dissociation energies for dipeptides of all common amino acid combinations are calculated in advance and stored in a database, and in the dissociation energy acquisition step, dissociation of the corresponding dipeptides from the database is performed when obtaining dissociation energy for any fragment ion of the candidate peptide. The peptide identification method using mass spectrometry according to claim 2, wherein energy is called and used. A peptide identification apparatus for determining an amino acid sequence of a peptide mixture in a test sample based on mass spectral data acquired by MS ⁿ analysis,
a) primary candidate selection means for selecting a plurality of candidate peptides based on mass spectral data;
b) For each of the plurality of candidate peptides, dissociation energy acquisition means for obtaining dissociation energy corresponding to all fragment ions;
c) Correlation for calculating a correlation coefficient between the dissociation energy obtained by the dissociation energy acquisition unit and the actually measured fragment ion intensity based on the mass spectral data estimated to correspond to each of the plurality of candidate peptides. Investigation means,
d) secondary candidate selection means for narrowing candidate peptides based on the correlation coefficient obtained for each candidate peptide by the correlation investigation means;
A peptide identification apparatus using mass spectrometry, comprising: Dissociation energy including the result of pre-calculating dissociation energy for dipeptides of all common amino acid combinations by modeling a peptide with 3 or more amino acids bound into a dipeptide consisting of 2 amino acids across the binding site of interest Have a database,
5. The mass according to claim 4, wherein the dissociation energy acquisition unit calls and uses the dissociation energy of the corresponding dipeptide from the dissociation energy database when obtaining the dissociation energy for an arbitrary fragment ion of the candidate peptide. Peptide identification device using analysis.

Images