Direct and automatic chromatography-mass spectrometry

Abstract

The invention discloses a method for acquiring sample spectrogram data, a spectrogram detection system and a computer readable medium. Detecting the presence of the compound in a time window; performing spectrum library search by using the spectrogram data from the time window; evaluating a hit list of compounds in each time window and selecting a subset of high likelihood compounds; carrying out regression analysis on the retention index of the compound in the subset and the measured retention time; abnormal values with too large retention index errors are identified and eliminated from the subsets; repeating regression analysis under the condition of removing abnormal values; calculating retention index values of all compounds in the whole sample operation; the calculated retention index value is compared with the retention index value of the possible compound to be identified, and the retention index matching degree is used as an additional index or a filter to further confirm the possibility of hitting each compound in the list.

Description

Technical Field

[0001] The present invention relates generally to the field of chromatographic separations associated with spectral detection systems such as gas chromatography (GC) with mass spectrometry (MS) detection, and more particularly to methods of acquiring, processing and analyzing the resulting separation and spectral data. Background Art

[0002] Ervin Kovats (Kováts E., Helv. Chim. Acta 41, 1915-1932 (1958)) introduced the use of the chromatographic retention time index, called the Kováts retention index (KRI or simply RI), a dimensionless number sometimes expressed as retention index units or simply "iu", as a method for converting chromatographic retention times (RT) to system-independent constants in gas chromatography (GC). The method depends on the calibration of the chromatograph, usually using a series of n-alkanes, where the Kováts index of an n-alkane is defined as 100 times its carbon number, for example n-hexane (C 6 H 14 ) is 600iu. Other organic compounds can then be assigned RI values ​​relative to these standards. Models have been established for both isothermal GC runs (logarithmic functions) and programmed temperature runs (linear interpolation functions).

[0003] If the RI of an unknown compound is available, it can be a powerful tool to help identify unknown organic compounds. The widely used NIST / EPA / NIH mass spectral library has updated its compound database to include measured RI values ​​for nearly 140,000 compounds (https: / / www.nist.gov / programs-projects / nist20-updates-nist-tandem-and-electron-ionization-spectral-libraries). In addition, NIST has developed an artificial intelligence (AI) model (e.g., see Matyushin, DD et al, Int. J. Mol. Sci. 22(17), 9194(2021) or Stein, SE et al, J. Chem. Inf. Model., 47(3), 975-980(2007)) that can calculate RI from chemical structure with high accuracy, providing RI values ​​for almost all of the more than 300,000 compounds in the ion spectrum library (EI-MS).

[0004] Although the RI value itself may not uniquely identify an unknown compound, when combined with a spectral library search in an EI-MS database (e.g., Stein, SE et al., J. Amer. Soc. Mass Spectrom. 5, 859-866 (1994)), a higher level of confidence in the identification of unknowns can be achieved than using any one technique alone. For example, a spectral library search can generally generate a list of the "best" possible matches between unknowns and compounds in the spectral library. However, this list usually does not provide a clear or unique qualitative result. Similarly, RI cannot provide a clear qualitative determination because many compounds may have similar RI values. However, the combination of the two can significantly improve the confidence in qualitative determination.

[0005] In order to correctly calculate the RI values ​​of unknowns, careful calibration with known compounds (usually an n-alkane ladder) must be performed under the same conditions as the sample run, including flow rate, temperature program, inlet temperature, column, etc. Since the RI values ​​are calculated by interpolation between the RI standard ladders, it is best to have the standard ladders evenly spaced and close to each other throughout the chromatographic run. If there are any changes in the method or conditions (flow rate, column length, etc.), the calibration ladder needs to be re-determined. In addition, over time, chromatographic conditions may change slightly due to ambient temperature, column degradation, or minor drifts in the GC electronics and flow controllers, which may require recalibration. Of course, the RI standards also need to be re-run when a new chromatographic method is used. Therefore, the disadvantage of this approach is that it requires additional time and effort to re-run calibration samples regularly, which adds additional burden and leads to reduced throughput and efficiency of the analysis. Another difficulty with using external standards (such as n-alkanes) is that some alkanes may not elute from the column before the GC programming is completed, resulting in carryover or other experimental complications, or the injected alkane standard mixture does not contain a high enough number of alkanes to cover the entire retention time range of interest.

[0006] Internal standards (known compounds spiked into the sample) can also be used to calibrate the GC. This is typically accomplished by spiking a compound with a known RI value into the sample. The advantages of this approach are that 1) the time interval between the standard and the analyte is minimized, thereby minimizing errors due to instrument drift, 2) the internal standard and sample are run under identical conditions to maximize accuracy, and 3) it eliminates the need to assay the standard separately, thus saving time and effort. However, many samples are inherently complex, and adding an internal standard to the sample can further complicate the analysis due to effects such as peak coelution or suppression during ionization, making it difficult to accurately determine the retention time (RT) of the internal standard. Additional steps are also required to ensure that the amount of RI internal standard added to the sample is comparable to the amount of analyte present in the sample itself.

[0007] Finally, the isothermal (logarithmic model) and temperature-programmed (linear interpolation model) RI models are applied to chromatographic separations with multi-step temperature ramps or temperature ramps combined with isothermal steps (see Figure 1 ) may lead to serious errors. Unfortunately, multi-stage temperature programs are common in GC because they can maximize separation power and minimize run time. Advanced RI models have been proposed in the literature to accurately calculate RI values ​​under such operating conditions, but it requires accurate measurement of the three thermodynamic constants ΔH, ΔS and ΔCp for each compound and careful confirmation of the key parameters of the gas chromatograph (Boswell, PG et al., J. Chromatogr. A. 1263: 179-88 (2012) and Peng B. et al., J. Chromatogr. A. 1374: 207-215 (2014)). Unfortunately, unlike traditional KRI, most organic compounds lack these thermodynamic constants. In order to accurately measure these constants, samples containing these compounds need to be measured under at least three completely different program temperature conditions and ensure that all of these compounds will peak within the time period of the run. This is a very challenging experiment and a difficult task, and therefore impractical.

[0008] Therefore, it is hoped that an effective method to overcome the above-mentioned defects and shortcomings of the prior art can be found. Summary of the invention

[0009] This application aims to improve the following:

[0010] 1. A new method for performing RI calibration without the prior use of any known standards, either added to the test sample itself or via external sample injection or run, saving time and effort while avoiding experimental, laboratory or human errors.

[0011] 2. These RI standards are self-generated by analyzing the test sample itself to select a portion of the analytes from the same sample as the RI standard.

[0012] 3. A method whereby analytes are picked from the test sample by spectral analysis and used as RI standards for RI calibration to ensure that exactly the same experimental or temperature programming conditions are used between the analytes in the standard and the test sample.

[0013] 4. A method for improving the RI standard used for RI calibration by iteration.

[0014] 5. Since RI calibration is performed internally on the current test sample itself, it is independent of the separation conditions.

[0015] 6. For each detected separation peak, the RI value calculated after calibration (fitting) is applied to further help identify unknowns. It can be used as an additional indicator outside the library search, or combined with the library search score to color-code the compounds in the compound list on the computer screen to indicate the possibility of correct identification, thereby greatly reducing the burden on analysts performing analytical tasks. This additional identification capability has greater credibility than identification using spectra or RI values ​​alone.

[0016] 7. Allows retrospective calculation of RI values ​​using old data that have been collected but do not have corresponding RI calibration data to improve the confidence in the identification of compounds that have already been analyzed.

[0017] 8. The RI values ​​thus obtained can be added back into an existing spectral library or database to supplement and enhance the collection of RI information, thereby continuously improving the ability to identify using spectral libraries.

[0018] 9. The prior art has shown that the accurate mass and spectral accuracy of a quadrupole mass spectrometer system can be calibrated to determine the elemental composition of a new unknown compound that is not yet included in a spectral library or database. Afterwards, the RI calculated from its RT can be further compared with the RI predicted by a computer model based on the elemental composition and the corresponding possible chemical structure to determine, select, rank or screen the possible chemical structures.

[0019] Each aspect is described below along with experimental results to illustrate their practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is an example of a temperature program for GC-MS analysis.

[0021] Figure 2 is a block diagram of a mass spectrometry system coupled to a separation device that can utilize the methods disclosed herein.

[0022] Figure 3 is the total ion chromatogram (TIC) of the GC / MS analysis of the thermally degraded condiment sample.

[0023] Figure 4 is the temperature program for GC-MS analysis of thermally degraded condiment samples.

[0024] Figure 5 This is the relationship between the RI value and the corresponding RT of the compound with the highest matching degree for each chromatographic peak of thermally degraded seasonings searched using the NIST spectrum database. Obviously, the correlation is very good, but relying solely on spectrum search will result in many outliers due to misidentification of compounds.

[0025] Figure 6This is a plot of RI vs. RT after eliminating all compounds that do not meet the minimum match requirement for the spectral search. The final regression line shown (in this case a polynomial) fits the remaining points very well, and the distribution of outliers shown is statistically significant. Points that are more than twice the standard deviation have been excluded from the regression analysis.

[0026] Figure 7 Screen shot showing GC / MS analysis of volatile organic compounds.

[0027] Figure 8 Included is a flow chart of one embodiment disclosed herein.

[0028] In multiple drawings, components or features that are identical are labeled with the same reference numerals. DETAILED DESCRIPTION

[0029] refer to Figure 2 , a block diagram of an analysis system 10 as described above, which can be used to analyze proteins or other molecules, and which references the characteristics of the present invention is shown. Although the present invention is described with reference to a single embodiment as shown in the figure, it should be appreciated that the present invention can be implemented in many alternative forms of embodiment. In addition, any suitable type of components can be used.

[0030] The analysis system 10 has a sample preparation portion 12, a mass spectrometer portion 14, a data analysis system 16, and a computer system 18. The sample preparation portion 12 may include a sample introduction unit 20, such as a Finnigan LCQ DecaXP Max manufactured by Thermo Electron Corporation of Waltham, MA, USA, for introducing a sample containing a molecule of interest into the system 10. The sample preparation portion 12 may also include an analyte separation unit 22, which is used to perform a preliminary separation of an analyte, such as a protein, to be analyzed by the system 10. The analyte separation unit 22 may be any of a chromatographic column for liquid or gas chromatographic separations, such as a gel separation unit manufactured by Bio-Rad Labolatories, Inc., Hercules, CA, which is well known in the art. Typically, a voltage or pH gradient is applied to the gel so that molecules such as proteins are separated as a function of one variable, such as migration rate (molecular weight, MW) through a capillary and isoelectric focusing point for one-dimensional separation (Hannesh, SM, Electrophoresis 21, 1202-1209 (2000)), or by more than one of these variables, such as by isoelectric focusing and by MW (two-dimensional separation). An example of the latter is known as SDS-PAGE.

[0031] The mass spectrometer section 14 can be a conventional mass spectrometer and can be any available mass spectrometer, but is preferably any of MALDI-TOF, quadrupole MS, ion trap MS, or FTICR-MS, or some combination, such as qTOF or triple stage quadrupole (TSQ). If it has a MALDI or electrospray ionization ion source, such an ion source can also provide sample input to the mass spectrometer section 14. Generally, the mass spectrometer section 14 may include an ion source 24; a mass spectrometer 26 for separating ions generated by the ion source 24 by mass-to-charge ratio (or simply mass); an ion detection section 28 for detecting ions from the mass spectrometer 26; and a vacuum system 30 for maintaining sufficient vacuum for the mass spectrometer section 14 to operate effectively. If the mass spectrometer section 14 is an ion mobility spectrometer, a vacuum system is generally not required and the data obtained is referred to as a mobility spectrum rather than a mass spectrum.

[0032] In parallel with the mass spectrometer part 14, there can also be other detection parts 23 for splitting the sample and detecting it almost simultaneously. Such other detection parts 23 can be a single-channel UV detector, a multi-channel UV detector, a refractive index detector, a stray light detector, or a radioactive detector, etc. Radioactive detectors widely used in experimental studies of drug metabolism with carbon-14 substitution can track metabolites in almost real time for comparison with mass spectrometry data.

[0033] The data analysis system 16 includes a data acquisition section 32, which may include one or a series of analog-to-digital converters (not shown) for converting the signal from the ion detector section 28 into digital data. This digital data is provided to a real-time data processing section 34, which processes the digital data by operations such as addition and / or averaging. A post-processing section 36 may be used to perform additional processing of the data from the real-time data processing section 34, including database searching, data storage, and data reporting.

[0034] Computer system 18 provides control of sample preparation section 12, mass spectrometer section 14, and data analysis system 16 in the manner described below. Computer system 18 may have a conventional computer monitor 40 to allow data entry and display of the results of the analyses performed on an appropriate display screen. Computer system 18 may be a computer system such as a computer system 18 that is configured to perform a quantitative analysis of the sample preparation section 12, mass spectrometer section 14, and data analysis system 16. or The computer system 18 is based on any suitable personal computer operating with an operating system or any other suitable operating system. The computer system 18 typically has a hard disk drive 42 on which the operating system and the programs for performing data analysis described below are stored. A drive 44 for receiving a CD or floppy disk is used to load the programs according to the present invention into the computer system 18. The programs for controlling the sample preparation part 12 and the mass spectrometer part 14 are typically downloaded as firmware for these parts of the system 10. The data analysis system 16 can be a program written in several programming languages ​​such as C++, JAVA or Visual Basic to implement the processing steps discussed below.

[0035] It is noteworthy that for the more general separation and spectral detection system coupling to which the present disclosure is applicable, the ion source section 24 can be replaced by a light source (for optical detection systems) or an energy source (for X-ray systems). The MS analyzer section 26 can be replaced by a dispersive device such as a grating (for optical systems with or without fluorescence options), and the ion detector section 28 can be replaced by an appropriate corresponding light or energy detector.

[0036] In a preferred embodiment, by Figure 2 The chromatographic / mass spectrometric system described in the present invention acquires samples and continuously acquires mass spectrometric data throughout the run, thereby generating the following Figure 3The run data shown is an example of a GC / MS run containing many chromatographic peaks. Chromatographic peaks identified with high matches by mass spectral library searches throughout the sample run are used as self-generated RI "standards" to calibrate the RI of the entire chromatographic run or any other chromatographic run, provided that these are run under substantially the same chromatographic or other separation conditions, including the stationary phase and carrier gas here. At least two standards can be used to identify the compounds used as RI calibration standards in a run. The first standard is the quality ("match") of the GC / MS spectral library search used to tentatively identify each peak. The match is typically calculated by comparing the electron impact (EI) ionization fragmentation pattern in the measured mass spectrum with pure standards available in an EI library (such as the library provided by NIST). The second is the relative retention time of other tentatively identified peaks located before and after the target calibration peak. This can be done in several ways, for example, by plotting the known RI values ​​of these compounds (available from the NIST database as previously obtained values) as a function of RT and performing a least squares fit or regression analysis on all tentatively identified compounds. This process can be used to select a subset of highly reliably identified and positionally important compounds as internal RI calibration standards to accurately calibrate the RI of all compounds in the entire chromatographic run. Once calibrated, the RI values ​​of all peaks in the chromatographic run can be determined and used to further improve the reliability of compound identification, rather than just searching the (NIST) library. The process flow is as follows:

[0037] a. Automatically detect or otherwise pick all chromatographic peaks in a chromatographic run and, if necessary, apply deconvolution of co-elutings to identify compounds contained therein, as described in U.S. Provisional Patent Application Serial No. 62 / 632,414 filed on February 19, 2018 and International Patent Application PCT / US2019 / 018568 published as WO2019161382.

[0038] b. Use the summed, averaged, or deconvoluted mass spectra within the RT window of the detected chromatographic peaks to search a spectral library (such as NIST / EPA / NIH or simply NIST) to generate a "hit list" sorted by match. Although the NIST spectral library is the most widely used public library, there are other similar or specialized spectral libraries that can be used, such as those used for forensics or flavors and fragrances. Homemade and proprietary spectral libraries can also be used. Other spectral libraries can also be used, such as the Sadler Infrared Spectral Database, which is widely used for FTIR spectroscopy analysis.

[0039] (http: / / www.ir-spectra.com / sadtler / sadtler.htm).

[0040] c. Evaluate each match in the "hit list" that is above a pre-specified "good" threshold (e.g., a NIST forward search score of over 900), and optionally set a limit on the difference between the forward and reverse search match scores (e.g., a difference of 30 or 50 in the NIST search match scores) to generate a subset of highly likely compounds. For coeluting compounds, if the detected chromatographic peak is a mixture of multiple compounds, the forward search match score may be low due to interference present in the spectrum, while the NIST reverse search match may be very high. This is because the reverse search masks or eliminates interference from irrelevant m / z values ​​in the spectrum. Therefore, the difference in search scores between the forward and reverse searches can be used as a good indicator of peak purity when selecting highly likely compounds.

[0041] d. Least squares regression is performed on the previously obtained RI and its RT of each identified high probability compound in the above subset. Note that multiple mathematical / statistical / numerical models can be used as regression models, including polynomials, spline functions, probability functions (such as logit functions), wavelets (such as Poisson wavelets) or appropriate combinations of these models, which can be performed over the entire RT range or for a given RT range. Certain graphical models or even pure numerical models (such as nearest neighbor clustering, replacement or averaging, etc.) can be used to replace a certain mathematical or statistical model mentioned above. It is also possible to iterate or exhaustively search among these possible models to derive the best model, which fits the existing data set well in the least squares or other sense, but will not overfit, that is, it has some residual errors, the order of magnitude of which is comparable to the known retention index measurement error limit of 5-20iu. When linearization is possible, such as in the case of polynomials, multiple linear regression (MLR) can be easily applied, using the methods cited in U.S. Patent Nos. 7,577,538 and 6,983,213.

[0042] e. Identify "outliers", i.e. compounds whose previously obtained RI values ​​differ significantly from the RI values ​​calculated from the regression model, e.g. based on one to two standard deviations as a cutoff or other statistical measure.

[0043] f. Remove outliers from the above subset used for regression.

[0044] g. Repeat the regression analysis for the remaining compounds in the subset until all compounds remaining in the subset are statistically consistent and a preset number of statistically allowed outliers remain.

[0045] h. For all compounds included in the entire run, their RI values ​​were calculated from their RT positions using the above regression model.

[0046] i. Compare the calculated RI value with the known RI value of the temporarily identified compound obtained by the library search, and use the obtained RI match as an additional indicator or filter to additionally evaluate the possible compounds retrieved from the library, or combine the RI match with other indicators (including spectral match) into a comprehensive overall match to more reliably rank the possible compounds to achieve a more reliable identification of the compounds. The combination of various match degrees can be in the form of an average of appropriately normalized or weighted values, or in the form of a sum of step (or S-shaped) functions, where some of the matches are only (significantly) taken into account if they are above a certain preset threshold.

[0047] The following diagrams show some examples of this process. Figure 3 Shown is the total ion chromatogram (TIC) from a GC / MS analysis of a thermally degraded food flavoring. The sample was complex, containing over 300 compounds as measured by the mass spectrometer. Figure 4 A temperature programming profile is shown, which contains multiple temperature ramps as well as isothermal portions at the beginning and end of the GC run.

[0048] Each peak picked or detected in a chromatographic run is searched against the NIST library to try to identify the corresponding compound. In the algorithm used by NIST, the match score represents the likelihood of a correct identification. For example, a match value above 900 is considered an excellent indicator of a correct identification. However, the highest match value may not necessarily correspond to the correct result due to minor changes in the spectral pattern caused by experimental or instrumental variations or possible structural / skeletal / positional isomers. However, the highest match value is more likely to be correct.

[0049] Next, the RI data for each compound with the highest match from the database can be plotted as a function of GC retention time, e.g. Figure 5 . Visually, it is easy to see a trend in the data but also to notice a large number of outliers. Outliers are either 1) compounds that were not correctly identified by the MS library search because they were not in the library, 2) peaks that were misidentified because they were a mixture of coeluting compounds, or 3) compounds that had a low signal-to-noise ratio in their spectra, resulting in a poor spectral match. Compounds corresponding to these outliers can be easily removed from the above plot by assessing the quality of the match using the spectral match score and / or the difference between the forward and reverse search match scores (an indication of peak purity).

[0050] Once these outliers are identified and removed, the RI values ​​for the remaining compounds can be plotted and fit to a mathematical / statistical function. As mentioned earlier, GC temperature programming with multiple temperature ramps can cause retention times to deviate in a nonlinear manner. In order to take into account such deviations from the classical logarithmic or linear Kovats model, a more flexible function can be used for fitting, such as Figure 6 The higher order polynomial shown.

[0051] However, the possibility remains that the match is not always correct. This is especially true for compounds with similar spectra, which may have the same chemical formula but different chemical structures (known in the art as various forms of isomers). It is well known that although these compounds are spectrally similar, their RI values ​​may differ significantly. These outliers can be easily identified as statistically significant deviations from the fitted curve. Figure 5 These outliers are easily identified in the plot shown and can be eliminated by their large deviations from the fitted curve.

[0052] Once outliers are detected, the process of fitting the data can be repeated to further remove outliers (if necessary). The number of iterations can be set by a reasonable statistical cutoff value, for example, 5% of statistical outliers can be allowed to be outside of twice the standard deviation (95% confidence interval under a normal distribution). The final fitted curve is the RI calibration curve, and the corresponding RI value can be calculated for each peak and compound in the chromatographic run based on the measured retention time.

[0053] Experimental comparisons found that the RI values ​​generated by this method are as accurate as the RI calibration values ​​generated by the more tedious and time-consuming external calibration (sequence) of n-alkanes using the traditional method.

[0054] Figure 7 Screen capture of a GC / MS analysis of volatile organic compounds is shown. Reported metrics include forward (NIST) and reverse (RevS) library match values, the difference between expected and measured RI values ​​(RIFit, where 100 indicates a perfect match and smaller values ​​indicate relative deviation from the expected value), and the spectral accuracy of the compound molecular ion (SA-M), as disclosed in U.S. Patent Nos. 6,983,213, 7,493,225, and 7,577,538 and U.S. Provisional Patent Application Serial No. 62 / 632,414 (formal application filed on February 19, 2018 and published as International Patent Application PCT / US2019 / 018568 as WO2019161382). Compounds that meet all metric threshold requirements are not color coded, while compounds with one or more metrics below the threshold are highlighted in red. This allows analysts to quickly and easily review data results and focus on "poor" compounds. Other intermediate color coding may be used to indicate "marginal" confidence in compound identification.

[0055] Figure 8The above steps in the flowchart of the first embodiment are shown, wherein at 51, spectral data is acquired during the separation of the analytes so that the separated peaks can be detected. At 52, a retention time window corresponding to the detected peak is selected, and the spectral data therein is summed or averaged to search a possible compound list (hit list) in the spectral library. At 53, high-quality or high-confidence compounds are selected from the hit list of all detected peaks in the entire chromatographic run to form a compound subset containing known RI values. These known RI values ​​are either stored together with the spectral library data or stored separately from the spectral library data in a separate database. At 54, a regression analysis is performed on the known RI values ​​of those compounds in the subset and their respective RT positions. At 55, the error between the fitted RI value of the regression model and the known RI value obtained from the spectral library search is calculated. At 56, compounds with RI errors greater than a preset or threshold value are marked. At 57, these compounds with large RI errors are then eliminated from the compound subset, and the regression is repeated until the RI errors of most (say 95%) of the compounds in the subset are within the allowable error. At 58, corresponding RI errors are calculated for all chromatographic peaks throughout the chromatographic run and used as an additional metric, or as part of an overall metric including a spectral search match score, to rank possible compounds for each peak detected, or to color code possible compounds to indicate the likelihood of correct identification.

[0056] Although the above description contains many details, these details should not be construed as limiting the scope of the invention, but should only provide illustrations of some possible embodiments. For example, due to the high reproducibility of a given GC / MS system, a regression model constructed from one sample run can be fully applied to another sample run under substantially the same separation conditions, or by adding additional high-probability compounds contained in another sample to an existing subset to dynamically enhance and improve regression models suitable for multiple runs or over time. In addition, the RI values ​​calculated from the sample run may have sufficiently good accuracy and can be added to an existing spectral library or database to replace missing or inaccurate data. In addition, sometimes the compounds from the new sample are true unknowns that are not yet included in the spectrum or database, then the measured spectral library data and the calculated RI values ​​can be added to the existing spectral library to enhance, expand or create a new spectrum or database. As described in U.S. Patent Nos. 7,577,538 and 6,983,213, there are certain advantages to acquiring spectral data in raw profile mode and calibrating the profile mode spectral data to obtain mass accuracy and spectral precision, which has certain advantages for creating, enhancing or utilizing accurate profile mode spectral data and databases, as disclosed in U.S. Provisional Patent Application Serial No. 62 / 830,832 filed on April 8, 2019 and U.S. Patent Application No. 16 / 843,505 (published as US2020-0232956 A1). Finally, even using a conventional quadrupole mass spectrometer, the elemental composition of an unknown compound that is not already included in the library can be determined, as described in U.S. Patent Nos. 7,577,538 and 6,983,213. For unknown compounds for which the elemental composition and RI value have been determined, chemical knowledge or databases such as ChemSpider (www.chemspider.com) can be used to search for possible chemical structures. Using the artificial intelligence (AI) models mentioned above (Matyushin, DD et al., Int. J. Mol. Sci. 22 (17), 9194 (2021) or Stein, SE et al., J. Chem. Inf. Model., 47 (3), 975-980 (2007)), the RI value for each possible chemical structure can be predicted, and then the predicted RI can be compared with the RI calculated from the sample run to determine the likelihood that a given structure is correct, thereby potentially providing a one-stop final answer to the compound identification problem through a one-time GC / MS experiment.

[0057] Therefore, the scope of the invention should be determined by the appended claims and their legal equivalents, rather than by the examples given. Although the content of the disclosed invention is described with reference to the described embodiments, it should be understood that it can be implemented with many alternative forms of embodiments. In addition, any suitable size, shape or part, element or material can be used. Therefore, this specification is intended to cover all forms of substitution, modification and variation within the scope of the appended claims.

[0058] It should be understood that the present invention may be embodied in a computer-readable non-transitory storage medium, and implemented by a computer system executing the methods and steps described herein through stored program instructions. Such storage medium may include any storage medium mentioned in the above description.

[0059] The techniques described herein are exemplary and should not be construed as implying any particular limitations on the disclosed invention. It should be understood that various substitutions, combinations, and modifications may be devised by those skilled in the art. For example, the steps associated with the process described herein may be performed in any order unless the steps themselves are otherwise specified or provided for. The disclosed invention is intended to cover all such substitutions, modifications, and variations within the scope of the appended claims.

[0060] The terms “comprise” or “comprising” should be interpreted as specifying the presence of stated features, integers, steps or components, but not excluding the presence of one or more other features, integers, steps or components or components.

Claims

1. A method for analyzing compounds by time separation using a spectral detection system, the method The following steps are involved: a. Obtaining spectral data of the sample; b. Selecting relevant retention time windows based on the presence of compounds of possible interest; c. using the spectrum data obtained within the relevant retention time window to search the spectrum library to identify possible compounds and their retention index values; d. using a regression model, using the identified subset of compounds, to perform a regression analysis between the retention index values ​​of the possible compounds and their retention times; and e. Calculate retention index values ​​based on the measured retention times of the compounds using the regression model.

2. The method according to claim 1, in, The separation technique is one of gas chromatography (GC), liquid chromatography (LC), supercritical fluid chromatography, ion chromatography (IC), capillary electrophoresis (CE), gel electrophoresis, ion migration and thermal lysis.

3. The method according to claim 1, in, The spectrum detection system is a sector mass spectrometer, a quadrupole mass spectrometer, a time-of-flight (TOF) mass spectrometer, an orbital trap mass spectrometer, a Fourier transform ion cyclotron resonance (FT ICR) mass spectrometer, and includes one of ultraviolet (UV), visible (Vis), near infrared, infrared, Raman, fluorescence, X-ray, and X-fluorescence spectra.

4. The method according to claim 1, in, The retention time includes one of chromatographic retention time, elution time, drift time and separation time.

5. The method according to claim 1, in, The retention index values ​​are previously obtained after calibration from measured retention times using gas chromatography standards related to n-alkanes.

6. The method according to claim 1, in, The regression model is one of a polynomial of a given order, a spline function of a given order, a probability function of a given form, a certain form of wavelet, a graphical model, a numerical model, and a combination of these models segmented within a given retention time window.

7. The method according to claim 1, in, The subset is selected based on library search masses above a given mass threshold.

8. The method according to claim 1, in, The subset is selected based on one of the differences or combinations between the forward and reverse library searches.

9. The method according to claim 1, in, The regression model was improved by iteratively eliminating from the subset those compounds whose calculated retention index values ​​were significantly different from those obtained previously.

10. The method according to claim 1, in, Improve the regression model by evaluating and selecting at least one of mathematical, statistical, graphical, and numerical models.

11. The method according to claim 1, in, The regression model is enhanced by incorporating additional compounds identified from other samples under substantially the same separation run conditions into the existing subset.

12. The method according to claim 1, in, The calculated retention index values ​​are compared with the retention index values ​​obtained previously for possible compounds identified by library searching.

13. The method according to claim 1, in, The difference between the calculated retention index value and the previously obtained retention index value is used as an additional metric or combined with other metrics to filter or rank possible compounds.

14. The method according to claim 13, in, One of the other metrics includes the library search match.

15. The method according to claim 1, in, The difference between the calculated retention index and the previously obtained retention index, or a comprehensive match that includes this difference, is used to color-code the compounds in the compound hit list, where a certain color corresponds to a certain probability of correct identification.

16. The method according to claim 1, in, The regression model is applied to the same sample and one of the subsequently injected samples under substantially the same separation conditions to calculate the retention index of the compound based on its retention time.

17. The method according to claim 1, in, The regression analysis may use multiple linear regression using one of matrix computation, matrix inversion, singular value decomposition, principal component analysis, and partial least squares.

18. The method according to claim 1, in, Add calculated retention index values ​​for new or additional compounds to an existing library or database.

19. The method according to claim 1, in, Spectral data for a subset of compounds are acquired in Profile mode and used to create or add to a Profile mode library.

20. The method according to claim 1, in, The acquired spectrum data has been calibrated for one of mass accuracy and spectrum accuracy, so that the elemental composition of a new unknown compound that is not included in the spectrum library can be determined.

21. The method according to claim 1, in, The calculated retention indices are compared to computer-predicted retention indices based on possible chemical structures for a given elemental composition to help identify, select, rank, and screen possible structures for compounds.

22. A spectrum detection system including a mass spectrometer, comprising a spectrometer operated according to the method of at least any one of claims 1 to 21.

23. For use with a computer in conjunction with a spectrometric detection system comprising a mass spectrometer, the computer readable medium having computer readable program instructions readable by the computer for causing the mass spectrometer to operate according to the method of at least one of claims 1-21.

Images