JP2005510732A - System and method for automatic protein sequencing by mass spectrometry - Google Patents

Abstract

A method for estimating protein sequences by analyzing data obtained by tandem mass spectrometry. The protein is subjected to partial isotope labeling by enzymatic digestion in an aqueous mixture containing non-natural abundance H ₂ ^1s O. Differential scanning mass spectrometry is applied to the peptide fragments obtained from this digestion process. Analyze the peaks in the spectrum to see if they originated from isotopically labeled fragments. Compute a filtered spectrum containing only peaks from y ions. The sequence of this peptide is deduced by calculating the mass difference between adjacent y-ion peaks.

Description

  The present invention relates generally to computer-implemented methods for determining the amino acid sequence of a protein by automatically interpreting the mass spectrum of the isotopically labeled C-terminal peptide fragment of the protein.

  The linear arrangement of amino acids in a protein is revealed by protein sequencing. Knowing protein sequences is essential for molecular biology techniques. For example, protein sequence information is a prerequisite for performing DNA cloning and provides information for creating oligonucleotide probes and PCR (polymerase chain reaction) primers. In addition, protein sequencing can synthesize peptides used to produce antibodies, identify proteins of interest, and help characterize recombinant products.

  If the sequence of a peptide sample is deduced without additional information, such as a known related peptide sequence, the approach is called de novo sequencing. Despite the development of genomic DNA sequencing, de novo sequencing of proteins and peptides in a biological research environment is still required. This is because many experiments are performed in vivo with unsequenced genomes.

  The basic method of protein sequencing is Edman degradation, a three-step chemical process based on N-terminal cleavage (Ward, Simpson, Proteins and Peptides, Isolation for Sequence Analysis of, Molecular Biology and Biotechnology, Robert Supervised by A. Meyers, VCH Publishers (1995), p. 767). Today, laboratory automation has made the implementation of the Edman method very efficient, but has several drawbacks including sensitivity to non-protein contaminants (e.g. Keen, Findlay, Protein Sequencing Techniques, Molecular Biology and Biotechnology, supervised by Robert A. Meyers, VCH Publishers, (1995)).

  Chemical sequencing of the C-terminus of the protein can be achieved by the thiocyanate method (Schlack, Kumpf, Physiol. Chem., (1926) 154, 125-170). Although useful for sequencing N-terminally blocked proteins and peptides, this method also has drawbacks, including the stringency of the reaction conditions and the need to attach the protein to a solid support (Bailey, J. Chromatog A, (1995) 705, pp. 47-65).

  Eventually, MS (mass spectrometry) has emerged as an attractive alternative to chemical methods and is used to solve sequencing problems that cannot be easily handled by the prior art of protein chemistry ( For example, Carr, Annan, “Overview of Peptide and Protein Analysis by Mass Spectrometry”, supervised by Current Protocols in Molecular Biology, Ausubel et al., John Wiley & Sons, (1997), 10.21). In mass spectrometry, the molecular weight of gas phase ions formed from intact neutral molecules is determined by separating them based on the m / z (mass to charge ratio) of the ions.

  One effective way to sequence proteins is to determine the molecular weight of a mixture of peptides, such as those resulting from proteolytic digestion, using mass spectrometry. For example, digestion of a protein with a specific enzyme such as trypsin cleaves the protein at a specific site. The position of this site is determined by the amino acid sequence of this protein. The result is a collection of peptides that give rise to a signature mass spectrum, often called a “fingerprint”. When the measurement accuracy of the m / z value is better than 0.01%, the amino acid composition of the peptide fragment can be estimated with high reliability. Therefore, the fingerprint can be used to clearly identify the protein, or it can be verified by comparing the translation product with information contained in the peptide fingerprint database of known proteins. it can.

  Mass spectrometry is not limited to the measurement of the mass of a single chemical species, and structural information including peptide sequences can also be revealed by MS / MS (tandem mass spectrometry) technology. In many mass spectrometry systems, gas phase ions are further fragmented either spontaneously or by collisions of gas molecules with so-called CID (“collision induced dissociation”). The generated subfragments can also be separated from each other by the m / z ratio.

  A unique advantage of tandem mass spectrometry is the ability to provide peptide amino acid sequence information at the picomolar or femtomole level. In this application, tandem mass spectrometry generally uses a first mass analyzer to screen for specific peptide ions, thereby subjecting those ions to fragmentation, for example with CID, to obtain the parent peptide or peptide fragment. Of subfragment ions can be generated. In this technique, a second mass analyzer is also used to first isolate and analyze the subfragment ions after peptide ionization and ion sorting. The resulting mass spectrum contains the m / z ratio of these subfragments.

  For example, the mechanism of fragmentation that organic molecules undergo in CID is well studied. Thus, by analyzing the masses of the parent species and their subfragments together, important structural information can be revealed. Specifically, molecules tend to cleave first from weak chemical bonds, but many functional groups remain intact during the fragmentation process. It has been found that the amide bond of the peptide is particularly susceptible to cleavage under the conditions used in MS / MS. As a result, the tandem mass spectrum of a peptide contains peaks corresponding to subfragments in which a single amino acid residue differs from one another, and thus can aid in sequencing (e.g., Hunt et al. (1986) Proc. Natl. Acad. Sci. USA, 83, 6233-6237).

  Nevertheless, problems with the analysis of tandem mass spectra remain cumbersome for several reasons. First, cleavage at the peptide bond yields a pair of fragments. One contains the N-terminus and the other contains the C-terminus. After fragmentation, it is unpredictable which of these fragments carries a charge, so most spectra contain a C-terminus known as the y ion (also called the Y "ion) It contains two series of fragments of the N-terminus, also known as b ions (also called B ions), the main challenge of de novo sequencing by mass spectrometry is to identify ions of one series of fragments, usually complex spectra It is to recognize with high reliability.

  Second, this fragmentation process is not ideal. Some amide bonds do not cleave during CID, so the difference between several peaks in the MS / MS spectrum corresponds to more than one residue, not the mass of a single amino acid residue To do. Similarly, certain types of fragmentation occur in amino acid residues, producing subfragments whose mass does not differ exactly from the mass of other subfragments by the number of amino acid residues.

  Third, multivalent ions often occur in conditions where the peptide sample is ionized. Thus, there can be a series of peaks in the spectrum that correspond to ions of the same fragment carrying different charges. In such a situation, the peak corresponding to a fragment where a single amino acid residue is different will differ by a fraction of the m / z value of the mass of this residue.

  Finally, another problem, which depends on the resolution of the instrument used, is that it may not be possible to resolve closely aligned peaks corresponding to differently substituted forms of the same fragment.

  Thus, in general, the spectrum obtained as a result of a typical de novo MS / MS analysis of a polypeptide is far from simple and straightforward to interpret, so that in general, some components of the peptide sequence are not identified.

  To date, computational methods for interpreting de novo MS / MS peptide spectra have been only partially successful. Part of this is that the spectrum itself does not have sufficient sensitivity or resolution to allow complete analysis. Another reason is that the algorithm used is too time consuming and impractical, or is not accurate enough and useful. For example, one initial approach uses the same cleavage specificity to compare the mass measurements of peptide fragments after enzymatic digestion with the theoretical peptide mass from each sequence entry in the database. This comparison yields a score that quantifies the goodness of agreement (e.g., Cottrell, Pept. Res., (1994), 7, 115-124, Matsui et al., Electrophoresis, (1997), 18, see pages 409-417).

  Another approach is to fit the theoretical spectrum of many possible sequences to the actual spectrum and continue this until a good match is obtained (Eng, JK et al., “An Approach to Correlate Tandem Mass Spectral Data of Peptides with Amino Acid Sequences in a Protein Database, Journal of the American Society of Mass Spectrometry, 5, 976-989, (1994)). The disadvantage of this approach is that there can be a huge number of combinations in connection with attempting to match the many possible sequences of amino acid residues into a set that produces a spectrum. The use of approximations to limit the vast number of such combinations that inevitably occur can often eliminate the correct sequence.

  In situations where the protein in question can be highly homologous to a protein with a known sequence, recently we have developed partial sequence data obtained by mass spectrometry and an efficient way to search large amounts of sequence databases. In combination, rapid protein identification was performed (Neubauer et al., Proc. Natl. Acad. Sci. USA, (1997), 94, 385-390, Neubauer et al., Nature Genetics, (1998), 20, 46-50). Similarly, direct analysis of large protein complexes was achieved by using computer algorithms to correlate the acquired peptide fragment mass spectrum with the predicted amino acid sequence in the translated genomic database. (Link et al., Nature Biotechnology, (1999), 17, 676-682).

  However, since the above technique relies on existing sequence data, it does not always affect de novo sequencing. The initial approach to de novo sequencing by mass spectrometry was to sequentially compare the mass difference between consecutive adjacent peaks in the spectrum with the amino acid mass until a match was found. Sequences were estimated based on scores related to peak intensities (Yates et al., `` Computer aided Interpreation of Low Enery MS / MS / Mass Spectra of Peptides '', Techniques In Protein Chemistry II, supervised by JJ Villafranca, (1991), Academic Press, page 477).

  Another technique for de novo sequencing is to assign a vertex to each peak and form a ridge between pairs of vertices that differ in mass by the mass of the amino acid residue, thereby measuring the so-called "spectral graph" from the measured spectrum. (Dancik et al., J. Comp. Biol., (1999), 6, 327-342). Only if the noise is efficiently removed from the spectrum can the correct sequence be inferred from the longest path of this graph. However, this method is based on the generation of a large number of sequence candidates, each with associated score probabilities, and the implementation of the inverse symmetric longest path problem, which is a theoretical technique based on graphs. I can not cope.

  Recent experimental techniques for sequencing proteins by mass spectrometry use labeling or tagging of peptide sequences. For example, in a chemical method in which the C-terminal residue is methyl-labeled by the formation of a methyl ester, the sequence data is derived from the characteristic peak intervals in the spectrum by comparing the spectra for the labeled and unlabeled samples. (Hunt et al., Proc. Natl. Acad. Sci. USA, (1986), 83, 6233-6237). The general disadvantage of chemical labeling is that it requires a chemical reaction step and that it is necessary to obtain spectra for two different samples.

  Another method by labeling replaces deuterium with acidic hydrogen along the peptide sequence (Sepetov et al., Rapid Commun. In Mass Spect. (1993) 7, pages 58-62). Although this method can quickly differentiate between peaks in the b and y ion series, this technique is only effective for short peptide sequences (less than 10 residues) and involves acidic side chains. Only additional sequence information about the residues is available.

  Prior to MS / MS analysis, isotopic labeling of peptide sequences with labels other than deuterium has been a desirable technique for some time, but requires sensitivity that is difficult to achieve with tandem mass spectrometry, and usually The spectra need to be compared for two samples. An example of a technique that can obtain information from a single spectrum is described in Gygi et al., “Quantitative analysis of complex protein combination using isotope-coded affinity tags”, Nature Biotech., 17, 994-999, (1999). ing. In this technique, proteins are labeled with a reagent such as iodoacetamide that has an affinity for sulfhydryl groups. The protein from one sample is labeled with a normal reagent and the protein from another sample is labeled with a reagent substituted with 8 deuterium. Both samples are combined and further labeled with a biotin affinity tag prior to analysis by mass spectrometry. The peaks from these two samples are separated by 8 mass units. The disadvantage of this method is that these protein samples require cysteine residues.

Although an application of ¹⁸ O labeling has been reported (Takao et al., Anal. Chem., (1993), 65, 2394) where two spectra are obtained for a single sample using a 4-sector tandem mass spectrometer. ˜2399), this analytical method is a simple comparison of the spectra and this method suddenly becomes impractical for sequences longer than those reported (about 10 residues). Therefore, de novo sequencing of proteins by mass spectrometry has been a challenging task for quite some time.

Recently, however, the sensitivity required to perform de novo sequencing of isotopically labeled peptides has been achieved by combining a nanoelectrospray ion source with a quadrupole time-of-flight tandem mass spectrometer. This approach takes advantage of the unique features of a quadrupole time-of-flight device that provides higher sensitivity and resolution than other types of mass analyzers (Shevchenko et al., Rapid Communications in Mass Spectrometry, (1997), 11, 1015 ~ 1024). For example, isotopically labeling C-terminal peptide fragments by enzymatic digestion of proteins in 1: 1 ¹⁶ O / ¹⁸ O water provides a characteristic and easily identifiable isotopic distribution for these fragments. (Schnolzer et al., Electrophoresis, (1996), 17, 945-953). The principle of this method is that they were obtained before tandem mass spectrometry, and when they were labeled 50% of the C-terminus with ¹⁸ O isotopes and 50% with ¹⁶ O isotopes Is to identify the C-terminal fragment ion of this peptide in one spectrum by its ¹⁶ O / ¹⁸ O isotope pattern. Two spectra are required, both from the same sample. Analyzing the difference between these two spectra, the subtraction result, means that measurements can be taken with enhanced sensitivity and a series of peaks from isotopically labeled subfragments are identified. These peaks arise from the C-terminal ions that differ in mass by one amino acid, from which this amino acid sequence can be revealed.

Nevertheless, analysis of isotopically labeled spectra remains complex for several reasons. Identification of C-terminal peptides with a characteristic isotope distribution, such as those obtained when digesting proteins in water with a known proportion of ¹⁸ O, is a natural isotope of isotopes such as ¹³ C and ¹⁵ N, among others It becomes difficult by the presence. For example, due to their natural isotope abundance, two peaks separated by 2 Da (Dalton) in the peptide mass spectrum may result in peptide subfragment ions having ¹⁶ or ¹⁸ O atoms at the C-terminus or It can result from peptide subfragment ions having ¹⁶ O atoms at the C-terminus, or one ¹³ C atom and one ¹⁵ N atom, or two ¹³ C atoms, or two ¹⁵ N atoms. As peptides get larger, the more likely ¹³ C and ¹⁵ N isotopes are not incorporated, the more difficult it becomes to identify C-terminal peaks and sequence amino acids Become.

  In summary, all existing methods for de novo protein sequencing have drawbacks. Mass spectrometry is a relatively promising technique for protein sequencing. This is because the amount of sample required can be picomolar or even femtomole and a very accurate spectrum can be obtained. However, comparatively large peptides and proteins have considerable spectral interpretation difficulties. Therefore, in the current technique, there is a demand for an analysis technique that can estimate the sequence of a large peptide from a mass spectrum.

  Citation of a reference herein shall not be construed as indicating that such reference is prior art to the present invention.

  The present invention involves deriving the amino acid residue sequence of a protein or peptide by automatically analyzing differential scanning mass spectrometry data. Specifically, an embodiment of peptide sequence analysis addressed by the present invention is to automatically identify C-terminal or y-ion peaks in mass spectrometry data. Once y-ion peaks are identified, the peptide sequence can be deduced by calculating the mass difference between adjacent y-ion peaks and considering each mass difference to be due to a specific amino acid residue. Since a peptide's mass spectrum consists of many peaks, it is usually not easy and rarely quick to derive a peptide sequence by human inspection of a simple difference between a pair of spectra. .

  The subject of the present invention is therefore a computer algorithm that estimates the peptide sequence of a peptide from a pair of MS / MS spectra obtained for a partially isotopically labeled sample. This algorithm seeks to compute a “filtered” spectrum that includes only the C-terminal set of subfragments (y ion series). The amino acid sequence can be accurately estimated from the result.

The present invention includes an instrument for determining the amino acid residue sequence of a peptide. The instrument was configured to receive mass spectrometry data obtained by applying differential scanning mass spectrometry to a peptide sample in which the isotope label is present at a rate that is substantially different from its natural abundance. An input device, a processor configured to perform mathematical operations on the mass spectrometry data, and a memory connected to the processor. This memory is executed repeatedly for each peak in the mass spectrometry data, and a first set of instructions instructing the processor to generate a probability that the peak in the mass spectrometry data is derived from the y-ion subfragment of the peptide; Instructs the processor to generate a filtered mass spectrum of the peptide, and each peak in the filtered mass spectrum having an intensity greater than a threshold corresponds to a y ion subfragment of the peptide A second set of instructions to predict and a third set of instructions to instruct the processor to derive the amino acid residue sequence of the peptide from the filtered mass spectrum and store it in memory Remember. In a preferred embodiment, the isotope label is ¹⁸ O and the proportion is 50%.

  According to the differential scanning mass spectrometry technique, the mass spectrometry data includes a first mass spectrum with signals from subfragment ions in the presence and absence of isotope labels, and subcarriers without isotope labels. And a second mass spectrum in which the signal from the fragment ions is substantially suppressed. In a preferred embodiment, the probability is calculated from the product of the first score value and the second score value. This first score value is derived from the first mass from an isotope cluster that includes a signal from a sub-fragment ion that does not have an isotope label and a signal from a sub-fragment ion that has an isotope label in that proportion. It is proportional to the likelihood that a peak in the spectrum will occur. The second score value is an isotope cluster that includes peaks from subfragment ions in which isotope labels are present in the above proportions, and peaks from subfragment ions that do not have isotope labels are in the first mass spectrum. Is proportional to the likelihood that a peak in the second mass spectrum will occur from an isotope cluster that is effectively suppressed compared to.

  The present invention further includes a method for determining the amino acid residue sequence of a peptide. The method includes receiving mass spectrometry data obtained by applying differential scanning mass spectrometry to a peptide sample in which an isotope label is present in a proportion substantially different from its natural abundance, and a first set. Is repeated for each peak in the mass spectrometry data to generate a probability that the peak in the mass spectrometry data is derived from the y-ion subfragment of the peptide, and the filtered mass spectrum of this peptide Generating and predicting each peak in the filtered mass spectrum that has an intensity greater than a threshold to correspond to the y-ion subfragment of the peptide, and from the filtered mass spectrum Deriving the amino acid residue sequence of the peptide. According to a preferred embodiment of the invention, the method for determining the amino acid residue sequence of a peptide is performed by a computer under the control of a program. The computer includes a memory for storing the program, an input device configured to receive mass spectrometry data, and a processor configured to perform mathematical operations on the mass spectrometry data.

  Other advantages of the present invention will become apparent upon reading the following detailed description and upon reference to the drawings in which:

Introduction Describes how to identify protein y-ion peaks in tandem mass spectra. As used herein, the term “protein” is used in a broad sense and includes peptides, polypeptides and oligopeptides, and derivatives thereof such as glycoproteins, lipoproteins and phosphoproteins, and metalloproteins. Will be modified as needed. A key feature that distinguishes such chemical species is that the “protein” contains one or more peptide bonds (—N (—H) C (═O) —).

The goal of this method is to simplify and automate the analysis of MS / MS spectra so that potential peptide sequences can be proposed. This method is implemented in the form of a computer algorithm. This algorithm obtains two peptide fragment ion spectra instead of just one from a protein sample that has been enzymatically digested in a water mixture containing known proportions of H ₂ ¹⁸ O and H ₂ ¹⁶ O. Based on. The water mixture, as part composition of H ₂ ¹⁸ O is substantially greater than the natural abundance of H ₂ ¹⁸ O, the conditions, the peptide fragment, at its C-terminus, the water mixture ¹⁸ O label is incorporated at the same rate as is present in By sorting and fragmenting the entire ¹⁶ O / ¹⁸ O isotopic mixture of this peptide, one spectrum is obtained, and a second spectrum is obtained in which only peptide ions labeled with ¹⁸ O are fragmented.

After obtaining the mass spectra of ¹⁶ O / ¹⁸ O and ¹⁸ O, the data is analyzed using the computer program products and methods of the present invention to identify peaks arising from y ions. When these two spectra are compared using two criteria, the peak corresponding to the C-terminal peptide subfragment can be identified. The first criterion is the ¹⁶ O / ¹⁸ O isotopic distribution of the C-terminal peptide subfragment in the first spectrum, which is usually difficult or impossible to clearly recognize by visual inspection. The second criterion is a change in the isotope distribution of the C-terminal subfragment ion peak when the first spectrum and the second spectrum are compared. The C-terminal ion is identified by including a peak from the complete ¹⁶ O / ¹⁸ O isotope distribution in the first spectrum but only a peak from the ¹⁸ O isotope in the second spectrum. Non-C-terminal ions have the same isotope representation in any spectrum. This is because neither spectrum contains ¹⁸ O isotopes at the rate introduced by enzymatic digestion.

  Once all C-terminal fragment ions have been identified, the peptide sequence can be deduced by calculating the mass difference between adjacent fragments and from the order of these fragments in the spectrum. The method and computer program product of the present invention may further include calculation of the subtracted mass spectrum and the filtered mass spectrum.

  The method of the present invention has a resolution that allows the machine to obtain a well resolved mass spectrum, especially for proteins or peptides of any length only if it can resolve different isotopes. Can be applied. The number of amino acids that can be read depends on the sequence, so for example, there are peptides that are 20 amino acids long, of which only 5 amino acids can be read, other peptides that are 25 amino acids long, Sometimes the whole group can be read. In general, as the peptide becomes longer than a certain size, the readable sequence tends to be shorter. Still, it turns out that peptides up to 3 kDa in size (about 30 residues) can be sequenced to a sufficient length (i.e. 20 of 30 amino acids can be read). Yes. There is no lower limit for sequencing.

Instrument The present invention comprises a system 100 for estimating peptide sequences from mass spectrometry data obtained from a mass analyzer 130, as shown in FIG. The system 100 includes a processor 102, generally a memory region 104 that will include both high-speed random access memory and non-volatile memory (such as one or more magnetic disk drives), a keyboard for entering user-specific parameters, Input device 106, which may include a mouse and / or touch screen display, output device 108 for printing or displaying a sequence of proteins or peptides, and at least one connecting processor 102, memory 104, input device 106 and output device 108 Two buses 110 are provided. Although not shown in FIG. 1, system 100 also preferably includes a network or other communication interface for communicating with other computers and other devices.

  Preferably, this memory includes an operating system 120 that provides basic system services, a file system 122, an analysis module 128 configured to analyze mass spectrometry data, a cache 126, and optionally a GUI (graphical). (User interface) 124 is stored.

  Preferably, the system 100 acquires mass spectrometry data from the mass analyzer 130 via the data channel 132. In one embodiment of the invention, the mass analyzer 130 is a triple quadrupole mass analyzer.

  When analysis module 128 receives a request to infer the sequence of a peptide or protein from mass spectrometry data, it has a sufficient probability that which peak in the mass spectrometry data corresponds to a peptide subfragment in the y ion series. Execute instructions that can be identified. Once the y ion is identified, the amino acid sequence is determined by calculating the mass difference between adjacent y ion peaks. Each mass difference corresponds to the mass of one amino acid residue. All amino acids in the peptide chain can be distinguished except for leucine and isoleucine, which have the same mass. Using the principles well known to those skilled in the art, the entire sequence of a protein can be determined by linking or overlapping separate peptide sequences determined from spectra of different peptide fragments.

  When operated in a laboratory environment with mass spectrometry data, this system can provide an efficient and useful method for deducing the amino acid residue sequence of a protein or peptide.

Apparatus Configuration The mass analyzer separates ions according to the m / z ratio, which is the ratio of their mass m to charge z. In the first stage, the sample is ionized, for example by electron impact, and in subsequent stages, ions are generated that are accelerated towards the detector by a non-homogeneous electromagnetic field. In one embodiment, this electromagnetic field causes the ion trajectory to perturb according to its m / z ratio. Ions with lower mass move faster and are less easily perturbed than heavier ions. Ions with a small charge number will perturb more than those with a large charge number. In fact, most of the ions produced carry only a single positive charge, but some ionization techniques can easily generate multivalent ions. For several reasons, different m / z ions are produced from the same sample. That is, molecules may dissociate depending on ionization conditions, ions themselves may later transfer and dissociate, and there are always many different isotope substituents in a given sample molecule. .

  In one embodiment, a triple quadrupole mass analyzer is used to acquire peptide subfragment data. An example of such a machine is API III from PE-Sciex (Perkin Elmer Sciex). In this embodiment, three quadrupoles are used as the ion guide, mass filter and collision cell. Although a typical arrangement of such a mass analyzer 300 is shown in FIG. 3, it should be understood that variations of such mass analyzer components are envisioned to be implemented with the method of the present invention. .

  Tandem mass spectrometry uses two-stage mass spectrometry. In the first stage, precursor ions are generated from the ionization source 304. In a preferred embodiment, electrospray ionization is used to generate these precursor ions. These precursor ions optionally pass through a first quadrupole 306 that serves as an ion guide. This ion guide is usually not a mass selective quadrupole and is usually only present in triple quadrupole machines. The precursor ions pass through a mass filter 310 that sorts out precursor ions having a specific value of m / z ratio, or more generally, m / z ratio that takes a narrow range of values. Currently, it is known that the mass filter 310 that provides maximum sensitivity is a quadrupole mass filter. Alternatively, an ion trap can be used as an alternative. In a preferred embodiment of the present invention, the mass filter 310 is a quadrupole mass filter. The range of m / z ratios that the quadrupole mass filter transmits is known as the transmission window.

  In a preferred embodiment, the mass analyzer 300 used to acquire peptide subfragment data is a “Q-TOF” (quadrupole time of flight) mass analyzer. An example of such a machine is the “Q-Tof2” from Micromass in the UK. Such a machine uses two quadrupoles. The quadrupole 312 is used as a mass filter for precursor ion sorting, and the quadrupole 322 is used in a collision cell 310 where the precursor ions are further fragmented into subfragments. A “TOF” (time of flight) mass analyzer 340 is used to examine these subfragment ions. FIG. 8 also shows a representative mass analyzer design for practicing the present invention.

Ionization techniques There are several ionization techniques that can be used to generate precursor ions for mass spectrometry analysis. These include, but are not limited to, electron ionization, chemical ionization, field ionization, field desorption, fast atom bombardment, plasma desorption, laser desorption, and electrospray ionization. It is not something. The two most commonly used ionization techniques for biomolecular analysis are “MALDI” (matrix-assisted laser desorption ionization) and “ESI” (electrospray ionization). The method of the present invention is independent of the ionization technique used.

  One embodiment of a mass analyzer for use with the present invention uses MALDI. MALDI is a specific type of laser desorption method in which a biomolecule is crystallized with a large number of excess small ultraviolet radiation absorbing organic acids (matrix). When the co-crystal is irradiated with an ultraviolet laser, the matrix molecules and biomolecules are in the gas phase, where protons move from the matrix molecules to the biomolecules to form biomolecule precursor ions for analysis. Usually, MALDI produces a monovalent precursor ion, which is then subjected to “PSD” (post-source decomposition) to produce fragment ions. Therefore, it is usually not necessary to use collision-induced dissociation with MALDI. Often, the MALDI method is used with a time-of-flight mass analyzer and can therefore be used with the method of the present invention.

  In a preferred embodiment for use with the present invention, the ionization source 304 generates precursor ions by ESI, and ions are accordingly formed by blowing a dilute solution of biomolecules from the tip of a fine metal capillary at atmospheric pressure. By this spraying, fine mist droplets are generated and charged strongly in a high electric field. As these droplets evaporate, the biomolecules receive one or more protons from the solvent and form one or more positively charged ions. As these droplets shrink, the repulsive force of the charges causes these ions to evaporate from the droplet surface and then analyzed in a mass analyzer. In a preferred embodiment of the mass analyzer used with the present invention, ESI is used to generate precursor ions. In MALDI, sample and precursor ions can be extensively fragmented, while in ESI, little or no fragmentation occurs. Furthermore, since the sample for ESI is in solution, this technique is ideal for supplying purification techniques such as HPLC.

Mass Filter In a preferred embodiment of a mass analyzer for use with the present invention, a quadrupole mass filter 310 is used to sort precursor ions. The quadrupole mass filter comprises a quadrupole 312 consisting of two pairs of precisely parallel metal rods, the opposing rods being electrically connected. A voltage consisting of “DC” (direct current potential) and an “RF” (AC high frequency) component are applied to each pair of rods. Ions passing through this quadrupole are alternately attracted to these rods and repel away from them, so these ions take a swing trajectory and only ions with a range of kinetic energy are Pass between the rods and exit on the other side. All other ions strike the rod. Since the kinetic energy of a given ion is proportional to the mass of that ion, the selection of ions is mass dependent. The ions passing through the quadrupole identify the quadrupole transmission window. If the DC and RF amplitudes are changed together while keeping their ratio DC / RF constant, the center of the transmission window can be shifted to other m / z values, and ions with different masses. Can be "filtered" through for analysis.

Generation of Subfragment Ions In one embodiment of a mass analyzer used with the present invention, filtered precursor ions having a specific m / z are sent to the collision cell 320. In a triple quadrupole mass analyzer, the collision cell includes a third quadrupole. In ToF machines, the collision cell typically includes the second of two quadrupoles. It should be understood that many machines that are compatible with the present invention use collision cells with quadrupoles. In machines that use ion traps, the ion trap itself is the collision cell. This is because ions can collide with the remaining gas atoms inside the ion trap.

  In the collision cell 320, the filtered precursor ions collide with uncharged gas molecules such as argon or xenon or dinitrogen fed from the source 314. The kinetic energy of these precursor ions is partially converted into vibrational energy, thereby breaking the main weak chemical bonds of the precursor ions. Peptide precursor ions are fragmented first from their peptide-amide bond to produce peptide subfragments. The resulting subfragment ions are analyzed by mass analyzer 340.

Mass Analyzer In a preferred embodiment, the mass analyzer used is a “TOF” (Time of Flight) mass analyzer. In this type of mass analyzer, subfragment ions are accelerated by an acceleration plate 342 and enter a region where there is no external electric field known as drift tube 344. If all subfragment ions entering this drift tube have the same kinetic energy given by 1/2 mv ² for ions of mass m and velocity v, the velocity is inversely proportional to the square root of the mass, so A subfragment 346 having a large mass will move more slowly than a subfragment 348 having a smaller mass. Thus, heavier subfragment ions have a slower time to reach the detector 350 at the end of the drift tube than lighter subfragment ions. TOF analyzers are often used with the MALDI method. TOF analyzers are advantageous in that they have virtually no mass range limitations and high scanning speeds.

  In the preferred embodiment, detector 350 is an electron multiplier that effectively displays the mass spectrum instantaneously. Detector 350 transmits mass data to computer system 100 via transmission channel 132.

  The limitation of TOF analyzers is that the peaks are broadened because not all members of the same subfragment ion assembly have the same kinetic energy. Since the initial energy spread is determined by mass, the peaks from heavier subfragment ions become wider. As is well known to those skilled in the art, the initial kinetic energy distribution of subfragment ions entering the drift tube can be reduced by increasing the final acceleration voltage. By increasing the length of the drift tube, the resolution of the TOF analyzer can be increased. This increases the arrival time difference between ions of different m / z, but also increases the arrival time spread of ions having the same m / z. In another embodiment of a mass analyzer for use with the present invention, the TOF analyzer is of the “reflectron” type in which ions follow a curved path. In a reflectron TOF analyzer, ions are slowed and redirected before they are directed to the detector. As ions change direction, relatively slow ions catch up with faster ions.

Mass Spectrometry Data Mass spectrometric data includes a plurality of elements, each element having an intensity value I for a certain m / z value. This data includes elements across a range of m / z values. A widely used unit name for the m / z value is Th (“Thomson”). A collection of data containing a range of intensity values expressed in Thomson is often referred to as a “mass spectrum”. In general, m / z values in a mass spectrum are separated from each other by 0.02Th, but can be separated from each other by 0.01Th or 0.05Th depending on the resolution.

  A “peak” in a mass spectrum is defined by a collection of adjacent elements, where each intensity value is greater than an intensity threshold. In general, mass spectrometry data also includes a number of low intensity data pieces, often referred to as background intensity and noise. This intensity threshold is selected so that noise is removed from the consideration during analysis. Usually, peak intensity is proportional to peak height, but this approximation may not hold for more complex spectra, especially for relatively heavy ions.

  Strictly speaking, the overall intensity of the peak is obtained by calculating the area under the peak. In one embodiment, this calculation is accomplished by a centroid method. In the center-of-gravity process, data in a window having a width of 0.08Th is added to an arbitrary peak having a width obtained by “FWHM” (full width at half maximum) of at least 0.04Th and added to this peak. In general, centroid processing is not fully satisfactory to obtain the accuracy required by the present invention because another peak can be erroneously synthesized. Therefore, in a preferred embodiment, the peak intensity is calculated using an integration method. Preferably, the method adds all intensities that exist around the peak within a window of about ± 0.02 Th. Different subfragment ions in the same spectrum may have different charge numbers, and different windows should be selected according to the charge number of the subfragment ions. Thus, for monovalent fragments, this window is preferably 0.04Th, and for divalent fragments, this window is preferably 0.02Th. Selecting another window when performing peak integration is compatible with the method of the present invention. In fact, it is possible to select different sized windows for different regions of the mass spectrum.

Most chemical elements that make up organic molecules have two or more isotopes that occur in nature. See Table 1 below. Since a mass spectrum consists of signals generated by a large number of ions, this spectrum involves statistical sampling of all naturally occurring isotopes. The larger the molecule, the greater the proportion of molecular assemblies that have one or more heavier isotope atoms. As a result, a given subfragment ion will not appear as a single sharp peak in the spectrum (unless an artificial isotope purity is obtained). Instead, the portion of the mass spectrum around the m / z value for a given ion contains multiple peaks. This is because each element present in this ion naturally has its own distribution of isotopes.

Thus, the mass spectrum of a given peptide subfragment will contain multiple closely separated peaks. Each of these peaks corresponds to a specific distribution of isotopes among the atoms of this peptide subfragment. If the peptide subfragment acquires a single charge during ionization, each adjacent separation peak for the subfragment is separated by approximately 1 m / z units. A cluster of peaks corresponding to fragments different from each other by the isotope deviation is called a cluster. There is no isotope with an integer mass number except ¹² C, which is defined to have a mass number of 12.0000 atomic mass units. Mass peptide molecule with one ¹³ C atoms, which involves one ¹⁷ O atom that is not the same as the exact mass of the same peptide molecule without the ¹³ C atoms. Therefore, the peaks in the cluster are not neatly resolved and may overlap significantly.

  The mass of a molecule that exists as an isotope in which all atoms are the most present is called “monoisotopic mass”. The monoisotopic mass of a molecule is the sum of the exact masses of the most abundant isotopes for all atoms. In general, the peak corresponding to this monoisotopic mass is of the minimum mass. This is because the most abundant isotope of each element present in a protein or peptide has the smallest mass among all isotopes. However, this peak is not always the strongest.

The intensity distribution of the peaks in the cluster is often referred to as the “envelope”, and its shape is the result of many contributing factors. For very large molecules, the peak corresponding to the monoisotopic mass does not necessarily have the maximum intensity. ¹³ C contributes the most to the isotope peak pattern of biomolecules. The presence of heavy isotopes of oxygen, nitrogen and sulfur also contribute to the isotope envelope. Carbon has two major isotopes that occur in nature. That is, the mass is 12.000000, the natural abundance is ¹² C of 98.9%, the mass is 13.003355, and the natural abundance is ¹³ % of 1.1%. Regardless of the size of the peptide, the first peak obtained by resolving the isotopic cluster arises from all ¹² C-containing ions. For peptides with a mass less than about 1,800 daltons (corresponding to peptides containing about 100 carbon atoms) this is the peak of maximum intensity. However, for peptides with a mass greater than about 1,800 daltons (corresponding to peptides containing more than about 100 carbon atoms), the first peak in this isotope cluster is not the peak of maximum intensity. This is because all ¹² C-containing ions are no longer the most abundant, ie, every molecule in this sample will contain on average at least one ¹³ C atom. In such cases, it may be more useful to consider the peak of maximum intensity and call it “average mass”.

Data Manipulation A feature of the present invention is to compare two mass spectra obtained from the same sample. These two spectra differ from each other by the center value of the transmission window. Simple subtraction rarely gives the difference between two spectra for the same sample, and several data processing operations should be performed. One problem is that subtraction can cause negative peaks due to phase mismatch. As an example, any of these two spectra may have a peak corresponding to an m / z value of 123. In the first spectrum, this peak starts at 122.85 and ends at 123.15, while in the second spectrum, the corresponding peak starts at 122.88 and ends at 123.18. These spectra need to be accurately aligned for subtraction to be effective. This is a procedure that can be difficult if the phase mismatch is not constant over the full range of these spectra. In practice, this problem is preferably addressed by “partial centroid processing”. The bin width is selected, which is generally 0.05Th and can be as low as 0.02Th depending on the mass resolution of the instrument. These spectra are separated into regular intervals of this bin width. If two data points are within this bin width, add their intensities.

The second complexity in spectral subtraction is that under slightly different operating conditions (which inevitably occurs in two different spectra), a pair of peaks common to both spectra do not necessarily have the same intensity. is there. Therefore, even when the phases are aligned, subtracting one peak from the other does not yield a baseline value, thus producing a small positive or negative peak. Although the magnification of these spectra can be changed to match the peak heights with each other, the required magnification can vary over the entire spectral range. In a preferred embodiment, the spectral magnification of the ¹⁸ O-containing peptide subfragment is altered and superimposed with the ¹⁶ O / ¹⁸ O spectrum. These spectra are typically divided into 20 Th wide windows. This width may be other values, both larger and smaller, and fits the method of the present invention. In each such window, the maximum peak in the ¹⁶ O / ¹⁸ O spectrum is determined, for example, m / z value m _p and intensity I1. The maximum peak is determined in the ¹⁸ O spectrum within the range of (m _p -1, m _p +2), and its intensity is I2. Change the magnification of the 20 Th width window of the ¹⁸ O spectrum by the coefficient I1 / I2. Finally, after subtracting the scaled spectrum from the unscaled spectrum, noise filtering is applied to the resulting spectrum. A peak having a width smaller than a certain threshold, for example, 0.05 Th is excluded.

Peptide Partial Isotope Labeling In a preferred embodiment, the computer program products and methods of the present invention are used in conjunction with partial isotope labeling of peptide fragments of proteins and differential scanning mass spectrometry techniques. Partial isotope labeling of the C-terminus of peptide fragments can be achieved by methods well known to those skilled in the art. A preferred embodiment for use with the present invention is shown in FIG. Peptides are labeled by enzymatic digestion of protein 200 in bulk solvent water, especially using trypsin, chymotrypsin or papain, preferably trypsin. A known proportion of this bulk solvent water is water labeled with ¹⁸ O, ie H ₂ ¹⁸ O (step 202).

This known proportion of labeled water is substantially different from the proportion of label found in nature. Preferably, substantially different means that when performing a mass spectrometric measurement, it is present in an amount that makes no contribution from the natural abundance isotopes, and is labeled from this labeled water. Is present in an amount that facilitates automatic analysis of the mass spectrum so that signals from peptides incorporating can be easily distinguished. In one embodiment of the present invention, the protein in the presence of labeled water 30% by volume of ¹⁸ O, preferably water were labeled with 33 vol% of the ¹⁸ O, more preferably ¹⁸ O 40 vol% Digestion is carried out in the presence of labeled water, most preferably water labeled with 50% by volume ¹⁸ O. In general, water labeled with a known proportion of ¹⁸ O from about 30% to about 75% by volume is suitable for carrying out the method of the invention. The volume ratio of water labeled with about 30% to about 75% ¹⁸ O is substantially different from the natural abundance of water labeled with ¹⁸ O.

One skilled in the art will appreciate that many peptide fragments are generated, for example, by enzymatic digestion of proteins in 50% H ₂ ¹⁸ O and 50% H ₂ ¹⁶ O. After digestion, these peptide fragments are purified and separated by, for example, gel electrophoresis or HPLC (step 204). The generated peptide fragment 206 is analyzed by mass spectrometry. Therefore, in the following, the term peptide will also include the term peptide fragment if it is understood that it is a peptide produced by fragmentation of certain relatively long peptides.

When an enzyme digests a protein, the enzyme cleaves the amide bond of the peptide, thereby corresponding to at least one peptide fragment with a free amino group (N-terminus) and an associated carbonyl group (C-terminus) Peptide fragments remain. Water molecules from the bulk solvent water are attached to the C-terminal group, producing carboxylic acid groups. Due to the presence of ¹⁸ O-labeled water of known proportions, cleaved peptide fragments of known proportion will have ¹⁸ O at the C-terminus. Preferably, the proportion of cleaved peptide fragments with ¹⁸ O at the C-terminus is approximately the same as the volume ratio of ¹⁸ O labeled water in bulk solvent water.

In a preferred embodiment, the known percentage of water labeled with ¹⁸ O is 50% by volume. For certain peptides digested in bulk solvent water, where 50% by volume is water labeled with ¹⁸ O, for each peptide fragment molecule of mass m with ¹⁶ O atoms incorporated at the C-terminus, ^{18 at the} C-terminus. There will be about one peptide fragment molecule with a mass of m + 2 since the O atom is incorporated. Thus, the peptide fragment containing the C-terminus and each sub-fragment of the peptide fragment will have a characteristic 1: 1 ¹⁶ O / ¹⁸ O isotopic distribution in the mass spectrum, which is only 2 mass units It should be distinguishable as two peaks of similar intensity separated. At a relatively low resolution, such a pair of peaks may appear as a single split peak. Therefore, the y ion in the M / S of such a sample should appear as a split peak, or a pair of similar intensity peaks. Unfortunately, in most cases, in the mass spectrum of the peptide fragment, 1: 1 ^16O / ¹⁸ O isotopic pattern visually distinguish things is impossible. Other ions may produce something similar to this peak pair, or split the distribution of peaks, and overlapping subfragment ions with similar masses may distort this pattern. For example, as already mentioned, it is sufficiently long peptide sequences for molecules with one ¹³ C substituent, peak at m + 1 is the peak of m corresponding to the absence of ¹³ C substituent molecule at least Just as big. If it is straightforward to identify the y-ion peak in this way, peptide sequencing by examining the mass spectrum of the isotope mixture would be feasible.

Herein, preferred embodiments of the present invention have been described using peptides labeled at the C-terminus with ¹⁶ O and ¹⁸ O utilizing trypsin digestion in H ₂ ¹⁶ O and H ₂ ¹⁸ O. However, it will be appreciated by those skilled in the art that the present invention can be used with peptide fragments that are partially labeled with other isotopes, such as ¹⁷ O, or that the present invention can use alternative methods of peptide labeling techniques. It will be understood. For other isotope labels, the amount of label to be incorporated may differ from the preferred amount for ¹⁸ O, but those skilled in the art will recognize that the amount of label different from the natural abundance to be used to practice the method of the invention It should be understood that can be determined.

Furthermore, in principle there are similar labeling schemes for b ions. In such embodiments, preferably the N-terminus is labeled. Isotopic labeling at the N-terminus is not as straightforward as isotope labeling at the C-terminus, which is easily realized at the same time as enzymatic digestion. ¹⁵ N-based signage is not ideal. This is because there are few practical reactions that can introduce such isotope labels into peptides. Therefore, preferably, labeling at the N-terminus is achieved artificially, for example, by acetylation. The acetylation reaction introduces a CH ₃ -C (= O)-group at the N-terminus (e.g. Pfeifer, T., Rucknagel, P., Kuellertz, G., Schierhorn, A., `` A strategy for rapid and efficient sequencing of Lys-C peptides by matrix-assisted laser desorption / ionisation time-of-flight mass spectrometry post-source decay '', Rapid Commun. Mass Spectrom., 13 (5), pages 362-9, (1999). ). When performing an acetylation reaction with a mixture of reagents, it should be possible to introduce a mixture of isotopes at the N-terminus, one containing the normal isotope and the other containing the heavier isotope. Since this acetylation reaction is an additional reaction carried out in this manner, isotope labeling is usually performed during the enzymatic digestion that is generally required to generate peptide fragments for sequencing. It cannot be realized with the same efficiency as in the case of C-terminal labeling. Further, preferably, component labeled with ^{_{isotopes, CH 3 -C (= 18 O}} ) - or ¹³ CH ₃ - ¹³ a C (= O) (both give mass shift of 2Da), these, More expensive than H ₂ ¹⁸ O, which can be easily purchased.

  The method of the present invention is not limited to sequencing peptide fragments obtained by enzymatic digestion of proteins. By the method of the present invention, the sequence of any peptide that has been subjected to partial isotope labeling can be determined.

Differential Scanning Mass Spectroscopy Differential scanning mass spectrometry provides two MS / MS spectra for a given peptide fragment 400 as outlined in FIG. ^A first spectrum denoted SP1 is obtained for a mixture of ¹⁶ O and ¹⁸ O containing peptides and their respective subfragments (step 402). ^{For only the 18} O-containing peptide and its subfragments, a second spectrum denoted SP2 is obtained (step 406). Thus, SP2 substantially suppresses signals related to ¹⁶ O-containing peptides and subfragments thereof. In a preferred embodiment, these two spectra are collected for the same peptide sample. The first and second spectra can be obtained in any order, but are separated by the step of re-centering the transmission window (step 404). Computational analysis of these two spectra (step 408) can produce a substantially clean spectrum for the C-terminal series of ¹⁶ O-containing peptide subfragments. The peaks that result from non-C-terminal subfragments always have their normal isotope distribution (regardless of ¹⁸ O labeling by enzymatic digestion), so when these two spectra are subtracted from each other these The peak should disappear. From this analysis, the peptide sequence 410 can be obtained.

  Peptide samples typically contain many different chemical species, for example, different peptide fragments resulting from enzymatic digestion, or specific peptides or peptide fragments in a form that is replaced by different isotopes. In a preferred embodiment, a particular peptide or peptide fragment in a form substituted by a different isotope constitutes a sample that is introduced into the mass analyzer. Thus, selection of one or more precursor ions by appropriately adjusting the transmission window allows analysis of a specific species or a limited subset of all of its species.

  The selection of precursor ions with a quadrupole mass filter inevitably compromises resolution and sensitivity. The resolution depends on the width of the transmission window. The narrowest window width provides the highest resolution, but the highest resolution requires the highest sensitivity. Therefore, if the quadrupole mass filter is operated so as to select a single isotope, the precursor ions are insufficiently transmitted and accurate analysis cannot be performed. That is, at the highest possible resolution, not enough sample is transmitted to obtain a useful spectrum at the sensitivity level used. However, the transmission window is not uniform. That is, ions having an m / z ratio that takes a value within the transmission window do not transmit with equal intensity. The method of changing the intensity over the entire transmission window of the mass filter is called a transmission curve.

The differential scanning mass spectrometry method of the present application states that due to the shape of the transmission curve of the quadrupole mass filter, the transmission window can be selected so that ions are effectively excluded and the sensitivity is not reduced accordingly. Based in part on people's unexpected discoveries. Without being bound by a particular theory, the shape of the transmission curve of a quadrupole mass filter is not symmetrical around the selected m / z, and the side that rises sharply (in the direction of decreasing m / z), and a relatively long tail (in the direction of increasing m / z). Because of this feature, recentering the center of a constant width transmission window, ie moving from one m / z to a slightly larger m / z, does not transmit smaller m / z. Therefore, by moving this window to a larger value, the lighter ¹⁶ O isotope does not enter this transmission window and the ¹⁸ O isotope enters it. This transmissive window behaves as having a sharp intercept “edge” at the lower edge of its m / z range.

The quadrupole mass filter makes it possible to select a transmission window corresponding to 3 Da, for example. When the transmission window is centered at an m / z value corresponding to the monoisotopic mass, the transmission window transmits both the ¹⁶ O and ¹⁸ O containing ions of a particular peptide, thereby obtaining the first spectrum SP1 . This transmission window is then re-centered near the second position of the m / z value corresponding to a value larger by 1 mass unit without changing its width, thereby reducing the signal-to-noise ratio. A second spectrum SP2 is obtained. At the second position of the transmission window, the transmission window effectively prevents the transmission of ¹⁶ O-containing ions without affecting the transmission of ¹⁸ O-containing ions. Thus, the transmission of peptides containing oxygen isotopes of lower molecular weight at the C-terminus is essentially completely suppressed in the second spectrum SP2. The second position of the transmission window allows ions having a mass 2 mass units greater than the monoisotopic mass to be transmitted. Such species include regular isotope variants of species that include ¹⁶ O (e.g., ions that include two ¹³ C atoms), but their contribution is not a natural proportion of ¹⁸ O due to enzymatic digestion. It is no longer important due to the contribution from peptide ions that incorporate In an alternative embodiment, the transmissive window can be centered at the second position before the first position.

  As shown in FIG. 2, the sorted precursor ions then enter the collision cell 320 and these precursor ions are fragmented into “subfragments”. As used herein, subfragments are also identified as “peptide subfragments” or “subfragment ions”. In the second stage of mass analysis, subfragment ions generated from precursor ions enter mass analyzer 340 and then to detector 350.

  In general, it is preferable to calibrate a mass analyzer in order to accurately assign mass from m / z values in the spectrum. As is well known to those skilled in the art, calibration can take the form of recording a spectrum for a sample whose mass is accurately known.

The 3Da transmission window is not so narrow that unacceptable loss of sensitivity occurs in the resulting spectrum. Thus, for a given fragment of a C-terminal peptide digested in 50% H ₂ ¹⁸ O and 50% H ₂ ¹⁶ O, the mass of the ¹⁶ O containing form is m, approximately the same intensity in the first spectrum Will result in only one peak in the second spectrum. These two peaks in the first spectrum SP1 correspond to fragments with masses m and m + 2, and a single peak in the second spectrum SP2 corresponds to fragment ions with mass m + 2. To do.

Mass resolution is often expressed as the ratio m / Δm. Where m and m + Δm are the masses of two adjacent peaks resolved in the mass spectrum and of approximately equal intensity. Differential scanning techniques require that mass analyzer 340 and detector 350 be able to resolve signals for subfragment ions that differ in molecular weight by no more than about 1 or 2 daltons. Specifically, a mass m peptide subfragment with a ¹⁶ O atom at the C-terminus and the same peptide subfragment with a mass m + 2 due to the C-terminal ¹⁸ O atom both have the same charge. However, the peaks arising from these two subfragments must be resolvable in the spectrum. The larger the peptide, the greater the mass of subfragment ions. Therefore, in order to be able to resolve the m and m + 2 peaks, the resolution of the analyzer must be higher for relatively large peptides.

  Precursor ions generated by electrospray ionization often have multiple charges, so that their m / z values are a fraction of their mass. A divalent subfragment ion appears to be half its mass at m / z values, but can resolve peaks for mass m and m + 2 subfragment ions separated by 1 m / z units. I will need it.

  One skilled in the art will appreciate that the resolution of the instrument used to collect the data will affect the accurate identification of the C-terminal ion. The method of the present invention can be performed on a low-resolution machine such as a triple quadrupole, but is preferably performed on a high-resolution machine.

All the C-terminal peptide subfragment ions are identified by the dual characteristic appearance of m and m + 2 in the ¹⁶ O / ¹⁸ O spectrum and by suppression of the corresponding ¹⁶ O peak in the ¹⁸ O spectrum. If identifiable, in principle, the peptide or protein sequence 216 can be “read” from this spectrum by noting the m / z difference between successive peaks in the C-terminal series. All amino acids except leucine and isoleucine, which have the same mass as each other, are distinguishable from each other by their characteristic mass and hence m / z value.

  However, in practice, peptide sequencing using differential scanning mass spectrometry is more difficult than simple spectral comparisons. Usually, identification of peaks arising from C-terminal peptide subfragment ions cannot be achieved by visual inspection, especially for relatively long peptides. The computer program product and method of the present invention alleviates this difficulty and allows for fast and accurate interpretation of mass spectra acquired using differential scanning techniques, thereby enabling previously unknown proteins. The amino acid sequence is determined quickly and accurately.

Algorithm for identification of C-terminal peptide subfragment ions The main problem addressed by the algorithm of the present invention is the identification of y ions in the mass spectrum of peptides. Its overall principle is to calculate the filtered spectrum SS for this peptide. See FIG. If m _p corresponds to the y ion of a ¹⁶ O-containing peptide, it is advantageous if this filtered spectrum is a simulated spectrum containing a peak with an m / z value m _p . The height of the peak in the filtered spectrum is similar to the intensity in the measured spectrum, but this peak is calculated by cumulative multiplication of a factor indicating the likelihood corresponding to the y ion . The advantage of a filtered spectrum is that it is visually satisfactory and easy to interpret.

Referring to FIG. 5, the steps prior to generating the filtered spectrum SS are as follows. It is preferable to confirm the number of charges of the peptide that gives rise to the spectrum. The starting points are the ¹⁶ O / ¹⁸ O mass spectrum SP1 500 and the ¹⁸ O spectrum SP2 502. From these, the number of charges of subfragment ions is estimated (step 504). Then, for each subfragment, analyze the peaks in the ¹⁶ O / ¹⁸ O mass spectrum to see if the peaks correspond to ions labeled with ¹⁸ O (step 506) and score for each peak. Estimate the value S1 508. ¹⁸ O peak in the mass spectrum of be analyzed, these peaks, ¹⁸ O ¹⁶ O / ¹⁸ O Mass spectrum ¹⁶ O-containing peptide subfragments present is suppressed as compared to in the mass spectrum of , Thereby generating a score value S2 512. It should be understood that the order of steps 506 and 510 can be reversed without departing from the scope of the present invention. Finally, the score values S1 and S2 are combined to generate a filtered spectrum SS514. Next, each step will be described in more detail.

This algorithm utilizes data on the ¹⁶ O / ¹⁸ O spectrum SP1 and the ¹⁸ O only spectrum SP2. The main task of this algorithm is to generate a score data set SD that assigns each peak in the ¹⁶ O / ¹⁸ O spectrum a probability value that the peak is the y ion of a ¹⁶ O-containing peptide subfragment. . Then, for each m _p values in accordance with equation (1) to calculate the filtered spectrum SS.

SS (m _p ) = SD (m _p ) * SP1 (m _p ) (1)
The net result of this algorithm is to generate a filtered spectrum that includes the calculated m / z value for ¹⁶ O y ions, excluding all other ions. It should be understood that the method of the present invention is equally applicable to the calculation of filtered spectra that correspond only to a portion or range of the measured spectrum. The method of the present invention should not be construed as limited to the computation of filtered spectra, score data sets or score values that include the entire spectrum measured for any position in the transmission window.

In conjunction with differential scanning mass spectrometry, the computer program products and methods of the present invention are used to facilitate identification of y ions by recognizing two essential features of y ions in the spectrum. First, the y ion has a ¹⁶ O / ¹⁸ O isotope distribution in the ¹⁶ O / ¹⁸ O spectrum SP1. Second, the ¹⁶ O peak of the y ion is suppressed in the ¹⁸ O spectrum SP2. Using these two features, an overall score value is calculated for each peak in the ¹⁶ O / ¹⁸ O spectrum, and that value forms part of the score data set SD.

The first step is to estimate the mass value m of fragments produces a peak at the position m _p. To achieve this, Uttenweiler-Joseph, S., Neubauer, G., Christoforidis, S., Zerial, M., Wilm, M., `` Automated de novo sequencing of proteins using the differential scanning technique '', Proteomics, 1 (5), 668-682, (2001). This is incorporated herein by reference. Depending on the type of ionization method used, the subfragment ions that give rise to a peak at m _p can be multiply charged. In general, the electrospray ionization method generates multivalent ions. Methods for estimating the number of charges are well known to those skilled in the art. The simplest and clearest way to identify multiply charged ions is to examine the spacing of peaks associated with adjacent isotypes. For example, if such peaks are separated by 0.5 m / z units, they are divalent ions. When these peaks are separated by 0.33 or 0.25 m / z units, they are trivalent ions or tetravalent ions, respectively. More sophisticated methods for interpreting the mass spectra of multiply charged ions include those described in Zhou US Pat. No. 5,072,115. This is incorporated herein by reference.

The overall score value SD (m _p ) for the peak at m _p gives the overall probability that this peak is the first peak of a doublet resulting from a partially labeled peptide subfragment. The score value SD (m _p ) is calculated from the product of the two factors in Equation (2).

SD (m _p ) = S1 (m _p ) * S2 (m _p ) (2)
S1 (m _p ) is the peak distribution in the envelope around the peak at m _p in the ¹⁶ O / ¹⁸ O spectrum and the intensity of the peak predicted for peptides of the same mass using natural isotopic abundance. The first score value, which is the probability calculated by comparing with the distribution and intensity. Thus, S1 (m _p ) has a peak at m _p in 50% H ₂ ¹⁸ O and 50% H ₂ ¹⁶ O, or in another embodiment, some other proportion of H ₂ ¹⁸ O. The probabilities arising from fragments of the ¹⁶ O / ¹⁸ O ratio obtained from enzymatic digestion in a water mixture containing are shown.

S2 (m _p ) compares the intensity of the peak at m _p in the ¹⁶ O / ¹⁸ O spectrum SP1 with the intensity of the peak at m _p in the ¹⁸ O spectrum SP2, and suppresses this peak in the second spectrum The second score value is the probability calculated by obtaining the degree of. S2 (m _p ) thus indicates the probability that the peak at m _p corresponds to the ¹⁶ O-containing y ion of the peptide.

Calculation of the first score value S1 based on the expected and observed isotope distribution In the first step in the method of the invention, a specific peak at position m _p is ¹⁶ O / ¹⁸ O in the spectrum SP1. Calculate the first probability arising from the first isotope of the isotope cluster. This first probability is known as the first score value. For peptides with a monoisotopic mass of m ₀ , the observed isotope envelope includes contributions from ions with masses of about m ₀ +1, m ₀ +2, m ₀ +3, etc. If these monoisotopic species cause peak intensity I ₀ at m _p, this envelope, each intensity I ₁ (m _p +1), the intensity I ₂ (m _p +2), intensity I It will include a continuous peak indicated by such as ₃ (m _p +3). If these ions that contribute to the cluster have a single charge, the continuous peaks in the envelope are separated by about 1 m / z units. The maximum mass that is usually considered depends on the value of m ₀ . This is because relatively large peptides are expected to incorporate a relatively large number of heavy isotopes and thus will have larger peaks in this isotope envelope. The observed values I ₀ , I ₁ , I _2, etc. of the peak intensity of this isotope envelope are usually determined by the natural abundance. The natural isotopic distribution of carbon, nitrogen, oxygen and sulfur (Table 1) was a factor that complicated the interpretation of peptide mass spectra, but it can be used somewhat advantageously in the present invention. Since the natural abundance is known, identify when this abundance is disturbed by an artificial ¹⁶ O / ¹⁸ O ratio resulting from, for example, enzymatic digestion in a mixture of 50% H ₂ ¹⁸ O and 50% H ₂ ¹⁶ O. Quantifying the degree to which this abundance is disturbed is straightforward.

The theoretical appearance of the peptide subfragment isotope envelope can be accurately modeled into a molecular ion cluster by solving a polynomial that calculates the total isotope of each element weighted by abundance ( Yergey, (1983), J. Mass Spectrom. Ion Process, 52, 337-349). Examples of equations used by this algorithm include, but are not limited to:

In these equations, a fragment of mass M _{pep has} a fragment of mass M _pep + n in its isotopic cluster. Intensity I _n fragment M _pep + n in the envelope can be calculated by adding the term [delta] _n. The initial value of all I _n is zero except for I _0. Since I ₀ is set to 1, no calculation is performed. I ₀ is the intensity of the monoisotopic species of this molecule. I _n is the intensity of the first and subsequent nth isotopes (including the (n + 1) th isotope).

To illustrate the application of these equations, consider the first and subsequent third isotopes. As discussed above, there can be more than one contribution to any isotope mass. There are contributions from ¹⁶ O-containing peptides with isotope substituents from other elements, so the contribution is calculated taking into account δ ₃ . However, there is also a contribution from ¹⁸ O-containing peptides. ^Since the peptide of ¹⁸ O is only 2 Da heavy, the first isotope after the first has the same mass in the peptide labeled with ¹⁸ O. Therefore, regarding the contribution of the peptide labeled with ¹⁸ O to the peak I ₃ , the contribution must be calculated considering δ ₁ . In this case, therefore, I ₃ increases by δ ₁ and δ ₃ .

  Examples of these formulas depend solely on the mass of the fragment and do not depend on element type, chemical composition or charge number. Therefore, before applying such an equation, the number of charges on the fragment must be estimated, whereby its mass can be determined. These equations are approximate equations obtained from the average abundance in currently known peptide sequences. Examples of the latest compilation of peptide sequences suitable for deriving these equations include, for example, “non-redundant” databases that are constantly updated by the EBI (European Bioinformatics Institute). For example, see http://www.ebi.ac.uk. This is also available from ftp://ftp.embl-heidelberg.de/pub/databases/nrdb. Another similar database compiled by NCB1 is available at http://ncbi.nlm.nih.gov/.

  Note that the heavier the peptide subfragment ions, the greater the intensity of the more peaks in the envelope. The envelope of a peptide subfragment with a monoisotopic mass of 1100 is, for example, the fragment whose intensity is calculated by the first equation and the mass is 1101, and the fragment whose intensity is calculated by the second equation and whose mass is 1102. You can see that For peptide subfragments with a mass of 1100, the contribution after 1103 is negligibly small.

In the differential scanning technique, the isotope envelope of the y ion is disturbed by the characteristic isotope distribution from partial labeling with ¹⁸ O. For example, the peak at m _p is the case of the first peak in ¹⁶ O / ¹⁸ O isotopic clusters of y ion peak (m _p +2), observations of strength such as (m _p +3), the It differs from the strength of naturally occurring subfragments of oxygen content. The characteristic double line of the ¹⁶ O / ¹⁸ O isotope cluster and the isotope envelope of the ¹⁸ O containing ion will be superimposed on the isotope envelope of the ¹⁶ O containing ion. In contrast, non-y ion isotopic envelopes simply follow the expected naturally occurring form. As a result, it is extremely difficult to visually assign peaks in the peptide mass spectrum resulting from y ions, but the isotopic envelope of these peaks is very different from the expected envelope for the naturally occurring isotopic distribution. Can be used in calculations.

The theoretically predicted peak intensities for ¹⁶ O-containing y ions based on natural isotope abundances can be calculated using polynomials of the type shown above, and they are calculated as I ₁ ^* , I ₂ ^* , I ₃ ^* etc. The observed and calculated intensity values can be normalized to I ₀ and I ₀ ^* , respectively, so that a quantitative comparison can be made. The score value S1 is a function of the difference between the observed value of the intensity for each peak in the isotope envelope and the calculated value, as shown in Equation 3.

However,

The absolute value delta _n of the difference in intensity, the peak (m _p + n) is an observation value of the intensity of I _n and a peak on the assumption that the peak occurs from ¹⁶ O-containing y ions in m _p (m _p Calculated from I _n ^* , the intensity calculated for + n). S1 _n (m _p ) in the formula (3) is a contribution from the peak intensity at (m _p + n) to the peak score value S 1 at m _p . The universal constant e (~ = 2.71828 ...) is the base of the natural logarithm.

Equation (3) has two parameters, which have the following effects. λ is “strength”. That is, the weight given to the score value, which can be adjusted according to how important this criterion is. σ is a parameter of the "sharpness", how fast S1 _n in accordance with an increase of delta _n affects or drops to zero. The form of equation (3) is shown in FIG. 6 (a). However, λ = 5 and σ = 0.25. Δ is a value on the x-axis, and S1 is a value on the y-axis.

The sharpness parameter σ determines how fast these score values fall to 0.001 * λ. Preferably, σ is not fixed but is calculated from the data itself. The purpose of the score function is to multiply the peak with ¹⁸ O isotope by the score strength λ and multiply the peak without this isotope by a very small value of 0.001 * λ according to the preferred form of equation (3). It is. Since most peaks have no ¹⁸ O isotope (only the C-terminal fragment has it), the average peak should be multiplied by a very small value close to 0.001 * λ. Therefore, σ is preferably selected so that the average peak is multiplied by about 0.003 * λ. That is, mathematically, it is as follows.

Where Δ _avg is the average of all values determined from this spectrum.

  In the preferred embodiment of the invention, λ is fixed at a value of 10.0. The values of σ and λ are affected by the machine used and the quality of the data. A person skilled in the art will be able to select better values for λ and σ than those given here, depending on the sample and machine used.

If the difference (Δ _n ) between the observed intensity value and the calculated value is large due to the exponent term in Equation (3), the value of S1 _n is reliably reduced when it is squared. Both spectra are normalized to I ₀ = I ₀ ^* = 1, so

Note that In another embodiment, the relative contributions of these two terms in Equation (3) are adjusted by changing the values of these coefficients separately while keeping their sum close to 1.0. can do.

  Other forms of equation (3) are possible without departing from the principles of the present invention.

  Note that the actual isotopic abundance of the fragments in the sample cannot also be determined completely. Two major reasons for this are that a given peak only contains signals from a small number of ions, so statistically the abundance of all isotopic substituents cannot be fully realized. , A given peak is often grown by signals from other ions, which results in unpredictable distortion.

For each cluster in the spectrum, depending on the mass of the peptide subfragment ion, the contribution S1 _n (m _p ) is calculated for the peak (m _p + n), thereby obtaining an isotopic envelope. For unlabeled fragments, consider only (m _p +1) if 300 <m ₀ <1000, consider peaks (m _p +1) and (m _p +2) if 1000 <m ₀ <1800 . Considering a potentially ¹⁸ O labeled fragment, the mass m ₀ +2 isotope is also considered. Therefore, when 300 <M _p <1000, peaks (m _p +1), (m _p +2) and (m _p +3) are included. It should be understood that the method of the present invention can also be implemented by calculating S1 across a subset of clusters in the spectrum.

Thus, the first score value S1 (m _p) is the observed value of the peak intensity in the isotope envelope around the given peak at m _p, a peak at m _p is ¹⁶ in SP1 O / ¹⁸ It is a measure of similarity to the intensity calculated for these peaks, assuming it is the first isotope within the O isotope cluster. This score value takes into account not only the degree of y ion labeling, but also the natural abundance of the isotope. In the past, this has complicated the mass spectrum of large peptides. A small difference between the observed intensity value and the calculated value is reflected in the form of a large score value. A large score value, indicating peak at m _p is, ¹⁶ O content y ions, i.e. the probability is for monoisotopic species is large.

In the last step, preferably, after calculating the S1 function for each peak m _p , the value of S1 is normalized to 1 by dividing the entire S1 function by its maximum value. This step effectively converts these score values into probabilities.

The second step in the method of calculating the present invention of the second score value based on the degree of suppression of the peak in the second spectrum, ¹⁶ O / ¹⁸ O isotopic clusters having ¹⁶ O isotope is suppressed in the spectrum SP2 Is to calculate the second probability that a specific peak at m _p occurs. This second probability is known as the second score value S2. This calculation is realized by comparing the two spectra SP1 and SP2, thereby determining the amount of peak suppression in SP2.

  As explained above, the transmission window of a quadrupole mass filter can be re-centered to a larger m / z value without reducing its width, which is effective for relatively light isotope transmission. Is disturbed. If the width of the transmission window is constant, the sensitivity is surely constant. In this way, two different spectra, SP1 from an isotopic mixture of a particular peptide and SP2 from a peptide containing only a heavy isotope, have similar signal-to-noise ratios.

The peak and intensity I ₀ at m _p in spectrum SP1 is collected when the quadrupole transmission window contains signals for fragments containing ¹⁶ O and ¹⁸ O, but due to this peak and intensity I _{0 at} m _p , An additional peak is generated, each having an intensity In and _denoted by (m _p + n). This is due to a mixture of naturally distributed isotopes and substituted fragments. Similarly, in the second spectrum in, quadrupole transmission window, the peak at m _p collected when it is centered around the fragment containing the ¹⁸ O isotope in the C-terminal has a strength indicated by K _0, peaks in the same envelope (m _p + n) has an intensity represented by K _n.

First, the intensity of the peak in SP1 resulting from the peak at m _p is normalized to I ₀ , and the intensity of the peak in SP2 arising from the peak at m _p is normalized to K ₀ . peak at m _p is, ¹⁶ O / ¹⁸ if O resulting from isotope first isotope in the cluster, this peak is suppressed in a second spectrum, the _{K 0 << I <SUB> 0} . By setting K ₀ = 1, the intensities K ₁ and K _{2 of} other peaks are artificially set to high values.

In calculating the second score value, it is desirable to average the abundant isotopes of fragment ions that should be able to be labeled with ¹⁸ O. For peptide fragments with m ₀ less than 1,400 Da, the peaks m _p and (m _p +2) are considered. If unlabeled, only the first isotope of m ₀ is present. When labeled, there are many isotopes m ₀ and m ₀ +2. For m ₀ greater than 1,400 Da, the peaks m _p , (m _p +1), (m _p +2) and (m _p +3) are considered. If not labeled, only the first isotope is present. ¹⁸ if it is labeled with O, all isotopes from m ₀ to m ₀ +4 there are many. This is because, as explained above, relatively heavy peptide ions increase the contribution of subfragments containing multiple isotopic substituents. The selection of 1,400 Da is not fixed, and other values can be selected within the region of about 1,400 Da without departing from the spirit of the present invention.

For each spectrum, the average relative intensity of the isotopes is calculated by averaging the intensities of all peaks considered. For a particular ion with mass m ₀ , these average values are I (ave) and K (ave) for spectra SP1 and SP2, respectively. Therefore, for m ₀ <1,400DA, I (AVE) = (I <SUB> 0 + I ₂ ) / 2, and for m ₀ > 1,400 Da, I (ave) = (I ₀ + I ₂ + I ₃ + I ₄ ) / 4.

peak at m _p is, if a first isotope in the ¹⁶ O / ¹⁸ O isotopic cluster, the peaks is suppressed in the second spectrum, the K (ave)> I (ave ). If peak at m _p is first isotope non y ions will K (ave) ~ = I ( ave).

An expression for the second score value S2 for obtaining the degree of suppression of the peak at m _p in the second spectrum is shown by Expression 6.

However, the parameter λ ₂ is the weight of the score processing given to S2 (m _p ), and Δ _n is the difference in peak intensity between these two spectra, that is, peak suppression. In the preferred embodiment, the scoring weight parameter is given the value 5. FIG. 6 (b) shows an example for λ ₂ = 5 and σ ₂ = 0.25. In FIG. 6 (b), Δ is a value on the x-axis, and S2 is a value on the y-axis.

If this suppression is negative, i.e. if the intensity of a given isotope in the second spectrum is greater than that in the first spectrum, then Δ is zero (S2 = 0, corresponding to no suppression). Set. As with the equation for S1, σ ₂ is a sharpness parameter. Since, by sigma _2, there is no suppression in the case (i.e., delta _n is small) is how fast the score value is from or fall to zero is determined.

This scoring function should have a value of λ ₂ for the peak labeled with ¹⁸ O. Such a peak is suppressed in the second spectrum. Since most of the peaks are not labeled, the average peak is not labeled with ¹⁸ O. For the average peak value, the score processing value should be very small, for example 0.002 * λ ₂ in the preferred embodiment. Furthermore, σ ₂ is preferably selected so that the average peak value is multiplied by a small coefficient such as 0.002 * λ ₂ . This mathematically means that σ can be expressed as a function of the form

Δ _avg is the average of all values determined from this spectrum, and β ₂ is a parameter that is preferably chosen to be 0.002. Of course, many other mathematical forms for σ are compatible with the method of the present invention.

If I (ave)> K (ave), the value zero is given to S2 (m _p ). This is not suppression of the peak at m _p in the second spectrum, therefore, this peak indicates that it is not a first peak in ¹⁶ O / ¹⁸ O isotopic clusters.

A large value of S2 (m _p ) indicates that there is a high probability that this peak will originate from y ions.

  In order to convert S2 values into probabilities, after calculating all S2 values, divide these score values by their maximum values.

Finally, the filtered spectrum SS is calculated using Equation 8 obtained by substituting Equation (2) into Equation (1) to calculate the intensity for the peak at each value of m _p be able to.

SS (m _p ) = SP1 (m _p ) * S1 (m _p ) * S2 (m _p ) (8)
The procedure described here is preferably repeated for each peak in both spectra or according to the number selected as the peak of interest.

As shown above, the score function for each peak is known for peak-specific parameters (such as suppression for ¹⁸ O-labeled peaks and deviations from the expected isotope distribution) and for all suppressions and all deviations. (That is, a parameter that can be calculated only when an average value Δ _avg is obtained). Thus, in the preferred embodiment, the calculation of the filtered spectrum begins with every deviation and suppression being calculated for each peak. The score values for all peaks are then calculated and the spectrum is then multiplied by these two score functions. Therefore, it is preferable not to omit any peak clusters. All calculations are performed for every peak, always evaluating each peak with respect to its characteristics, whether that peak can be the first peak of a ¹⁶ O / ¹⁸ O cluster. Even if one peak should be able to be the first isotope, it is premature to omit all peaks belonging to this cluster. This is because what has already been evaluated as the first peak cannot be speculatively determined to be the first peak before evaluating other peaks for the same purpose. For example, the second or third peak can be the first peak of a much more suitable ¹⁶ O / ¹⁸ O cluster.

Determining the amino acid sequence Once a series of y ions in the mass spectrum of a peptide has been identified, the sequence of this peptide can be deduced by examining the mass difference between adjacent y ion peaks in the spectrum become.

  The differential scanning method along with the algorithm of the present invention can identify peaks in the mass spectrum corresponding to this series of y ion subfragments. Each ion in this series contains the C-terminus of this peptide. Under ideal conditions of collision-induced dissociation that cleaves first from the peptide amide bond, each y ion corresponds to a peptide subfragment containing the exact number of amino acid residues. Thus, when the amide bond of each peptide is cleaved in the collision chamber, each y ion in this series differs in mass by the mass of the amino acid residue from the closest y ion. Except for leucine and isoleucine with the same mass, all amino acid residues have a unique mass, so lighter fragments are generated from relatively heavy fragments by calculating the mass difference between adjacent y-ion peaks It is possible to identify the amino acid residue itself that has been cleaved to show that it should be leucine or isoleucine.

  In one embodiment of the present invention, once the mass difference is calculated for a pair of adjacent y-ion peaks, the mass difference is sequentially compared with the mass of each of the 20 naturally occurring amino acid residues and matched. Do this comparison until you find what you want. If this mass difference is the same as the mass of one of these amino acid residues, this amino acid residue is assigned a corresponding position in the peptide sequence. Repeat this procedure for each adjacent pair of y ion peaks in the mass spectrum. In a preferred embodiment of the present invention, if the mass difference between two adjacent y-ion peaks does not correspond to the mass of one of the 20 naturally occurring amino acids, this mass difference is calculated for all amino acid residues. Search for a match compared to the sum of the masses of the pair. If one is found that matches the mass of a pair of amino acids, these two amino acid residues are placed in this sequence. The peptide-amide bond between this pair of amino acids was not cleaved easily enough in the collision chamber to produce another subfragment containing each residue of this pair. In this case, for example, it is impossible to infer the order in which this pair of amino acids is formed unless other information about other overlapping fragments is available.

  In a preferred embodiment of the invention, the procedure for matching the mass difference between adjacent y ion peaks is repeated for each distinct peptide or peptide fragment produced by enzymatic digestion of the protein. The sequence of each peptide or peptide fragment is deduced and the sequence of this protein is deduced by combining or overlaying the sequences of each fragment according to methods well known to those skilled in the art (e.g., Mann, M., `` A shortcut to interesting human genes: peptide sequence tags, expressed-sequence tags and computers ", Trends in Biological Science, (1996) 21, pp. 494-495).

Cited references All references cited herein are to the same extent as if each individual publication or patent or patent application was shown to be specifically and individually incorporated by reference in its entirety. Incorporated herein in their entirety by reference.

  Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, certain changes and modifications may be made to the invention without departing from the spirit or scope of the appended claims. However, it will be readily apparent to those skilled in the art in light of the teachings of the present invention.

  It will be apparent to those skilled in the art that many modifications and variations can be made in the present invention without departing from the spirit and scope of the invention. The specific embodiments described herein are provided by way of illustration only and the present invention is intended to cover the full scope of equivalents to which such claims are entitled. It shall be limited only by.

Alternative Embodiments The present invention can be implemented as a computer program product that includes a computer program procedure embedded in a computer readable storage medium. For example, the computer program product could include multiple separate program modules that could be stored on a CD-ROM, magnetic disk storage product or any other computer readable data or program storage product. These software modules in a computer program product are distributed electronically by transmitting computer data signals (embedded with these software modules) on a carrier wave over the Internet or other means You can also.

  Although the invention has been described with reference to several specific embodiments, the description is illustrative of the invention and is not to be construed as limiting the invention. Various modifications will occur to those skilled in the art without departing from the true spirit and scope of the invention as defined by the appended claims.

FIG. 2 illustrates a computer system according to the present invention. FIG. 2 is a diagram showing a quadrupole time-of-flight mass analyzer used in a preferred embodiment of the present invention. 4 is a flowchart illustrating a partial isotope labeling process used with a preferred embodiment of the present invention. It is a flowchart which shows the differential scanning method. 3 is a flowchart illustrating an algorithm according to the present invention. 6A is a diagram showing a typical shape of the score values S1 _n as a function of the calculated delta _n using Equation 3, FIG. 6B, the score as a function of delta _n calculated using Equation 6 It is a figure which shows the typical form of value S2 _n . FIG. 6 is a spectrum showing a comparison between an unfiltered peptide / subfragment / ion mass spectrum and a filtered peptide / subfragment / ion mass spectrum. 1 is a diagram illustrating a representative mass analyzer embodying the present invention.

[Sequence Listing]

Claims (26)

An input device configured to receive mass spectrometry data obtained by applying differential scanning mass spectrometry to a sample of the peptide in which an isotope label is present at a different rate than its natural abundance;
A processor configured to perform mathematical operations on the mass spectrometry data;
Connected to the processor and executed repeatedly for each peak in the mass spectrometry data, instructing the processor to generate a probability that the peak in the mass spectrometry data is derived from a y-ion subfragment of the peptide A first set of instructions, instructing the processor to generate a filtered mass spectrum of the peptide, each peak having an intensity greater than a threshold in the filtered mass spectrum; and A second set of instructions that predict to correspond to y ion subfragments and the amino acid residue sequence of the peptide from the filtered mass spectrum and store the sequence in the memory An amino acid residue of the peptide comprising: a memory for storing a third set of instructions for instructing the processor Apparatus for determining the column.   The mass spectrometry data includes a first mass spectrum having a signal from a subfragment ion in which the isotope label is present and a signal from a subfragment ion in which the isotope label is not present; and 2. The apparatus of claim 1, comprising a second mass spectrum in which signals from non-existent subfragment ions are substantially suppressed. The apparatus of claim 1, wherein the isotope label is ¹⁸ O.   4. The device according to claim 3, wherein the proportion is 50%.   4. The apparatus of claim 3, wherein the percentage is 33%. The probability is calculated from the product of the first score value S1 and the second score value S2,
(i) the likelihood that a peak in the first mass spectrum will occur from an isotope cluster containing a signal from a subfragment ion where the isotope label is not present, and (ii) the isotope label is present in the proportion The first score value is proportional to the signal from the subfragment ions,
The isotope label includes a peak from a subfragment ion present in the proportion, and a peak from a subfragment ion without the isotopic label is effectively suppressed compared to the first mass spectrum. 3. The apparatus according to claim 2, wherein the second score value is proportional to the likelihood that a peak in the second mass spectrum is generated from the isotope cluster. Wherein the first score value S1 is the peak at m _p is calculated from the following equation,

Where λ and σ are user defined parameters,

However, I _n is the n th peak intensity in the first isotopic cluster peak m _p occurs in the spectrum, I _n ^* is a peak m _p in the spectrum due to the natural isotopic abundance 7. The apparatus of claim 6, which is a calculated value of the intensity of the nth peak in the resulting isotope cluster.   8. The apparatus according to claim 7, wherein the value of λ is 10.0. σ is

The device of claim 7, wherein The second score value S2 is the peak at m _p is calculated from the following equation,

Where λ ₂ and σ ₂ are user-defined parameters,

However, I _n is the n th peak intensity in the first isotopic cluster peak m _p occurs in the spectrum, I _n ^* is a peak m _p in the spectrum due to the natural isotopic abundance 8. The apparatus of claim 7, which is a calculated value of the intensity of the nth peak in the resulting isotope cluster. The apparatus of claim 10, wherein the value of λ ₂ is 5. σ ₂ has β ₂ as a parameter

The device of claim 10, wherein   The third set of instructions calculates the mass difference between a pair of peaks in the filtered mass spectrum and compares the mass difference with the mass of each of the 20 naturally occurring amino acids. The apparatus of claim 2 comprising instructions. Receiving mass spectrometry data obtained by applying differential scanning mass spectrometry to a sample of said peptide in which an isotope label is present at a different rate than its natural abundance;
The first set of instructions is repeatedly executed for each peak in the mass spectrometry data to generate a probability that a peak in the mass spectrometry data is obtained from a y-ion subfragment of the peptide;
Generating a filtered mass spectrum of the peptide, and predicting that each peak having an intensity greater than a threshold in the filtered mass spectrum corresponds to a y-ion subfragment of the peptide; ,
Deriving the amino acid residue sequence of the peptide from the filtered mass spectrum.   The mass spectrometry data includes a first mass spectrum having a signal from a subfragment ion when the isotope label is present and a signal from a subfragment ion when the isotope label is not present; 15. A method according to claim 14, comprising a second mass spectrum in which signals from subfragment ions without an isotopic label are greatly suppressed. The isotope label is a ¹⁸ O, The method of claim 15.   16. The method of claim 15, wherein the percentage is 50%.   16. The method of claim 15, wherein the percentage is 33%. The probability is calculated from the product of the first score value S1 and the second score value S2,
(i) the likelihood that a peak in the first mass spectrum will occur from an isotope cluster containing a signal from a subfragment ion where the isotope label is not present, and (ii) the isotope label is present in the proportion The first score value is proportional to the signal from the subfragment ions,
The isotope label includes a peak from a subfragment ion present in the proportion, and a peak from a subfragment ion without the isotopic label is effectively suppressed compared to the first mass spectrum. 16. The method of claim 15, wherein the second score value is proportional to the likelihood that a peak in the second mass spectrum will occur from an isotope cluster. The first coefficient S1 is calculated from the following relational expression for the peak m _p :

Where λ and σ are user defined parameters,

However, I _n, said first spectral peak m _p in an n-th peak intensity of the isotope in the cluster occurring, I _n ^* is a peak in the spectrum which is calculated according to the natural isotopic abundance m _p a n-th peak intensity in the isotopic clusters occurring the method of claim 19.   21. The method of claim 20, wherein the value of [lambda] is 10.0. The second score value S2 is the peak at m _p is calculated from the following equation,

Where λ ₂ and σ ₂ are user-defined parameters,

However, I _n is the n th peak intensity in the first isotopic cluster peak m _p occurs in the spectrum, I _n ^* is a peak m _p in the spectrum due to the natural isotopic abundance 21. The method of claim 20, which is a calculated value of the intensity of the nth peak in the resulting isotope cluster. 23. The method of claim 22, wherein the value of [lambda] ₂ is 5. σ ₂ has β ₂ as a parameter

23. The method of claim 22, wherein   Generating a filtered mass spectrum for the peptide, calculating a mass difference between a pair of peaks in the filtered mass spectrum; and calculating the mass difference between 20 naturally occurring 16. The method of claim 15, comprising comparing the mass of each residue of the amino acid.   Executed by a computer under the control of a program, wherein the computer performs a mathematical operation on the mass spectrometry data, a memory storing the program, an input device configured to receive mass spectrometry data 26. A method according to any one of claims 14 to 25, comprising: a processor configured for:

Images