Multichannel detection and resolution of chromatographic peaks
First Claim
1. A method for resolving overlapped migration data obtained while performing qualitative analysis of a substance whose composition is to be determined, said method comprising the steps of:
- (a) subjecting said substance to be analysed to a first apparatus adapted to separate compounds contained in said substance as a function of their respective migration time in a given medium, said first apparatus further adapted to provide a set of migration data indicating the migration behavior of said compounds as a function of time;
(b) subjecting samples of said substance at specified times relative to the initiating of step (a) to a second apparatus, said second apparatus being adapted to measure and provide a unique set of characterizing time-invariant data for each of said samples, said set of characterizing data being a function of all the compounds contained in said sample at said specified times;
(c) identifying regions of peak overlap within said set of migration data, said regions of overlap indicating the presence within said sample of at least two compounds having approximately the same differential migration times;
(d) estimating a data set describing the differential migration time behavior of each of the compounds suspected of being present within each region of overlap identified in step (c), said set of estimated time behavior data describing the estimated shape of the migrational response of each said compounds within said region of overlap;
(e) calculating a set of identifying time-invariant data for each of said compounds suspected of being present within said regions of overlap from said measured sets of characterizing data and said estimated sets of time behavior data;
(f) varifying the goodness of the estimate made in step (d) by;
(i) combining the calculated sets of identifying data derived in step (e) with the corresponding sets of estimated time behavior data made in step (d),(ii) comparing said combination of said estimated and calculated data to said measured set of characterizing data obtained in step (b), and(iii) determining an error factor based on said comparison;
(g) modifying the estimated time behavior data made in step (d) as a function of the magnitude of the error factor determined in step (f);
(h) repeating steps (e), (f), and (g) as many times as is necessary in order to reduce the error factor calculated in step (f)(iii) to an acceptable level; and
(i) identifying the compounds present within said region of overlap based upon the last set of calculated identifying data obtained in step (e) as a result of the repetition of step (h).
2 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for resolving overlapped chromatographic peaks or other overlapped data obtained from operation of a differential migration apparatus used to analyze a substance whose composition is to be determined. The invention involves using a second analytical apparatus, such as a spectrometer, to supplement the data obtained from the basic differential migration apparatus. A multi-channel detector is used to monitor and record the various sets of data obtained from these apparatus. Data corresponding to regions of overlap identified by the apparatus are assembled in a measured data set. An estimated data set is then compiled describing the estimated migrational time behavior of each of a plurality of compounds present within the overlapped region. An identifying data set is then calculated based upon the estimated data set and the measured data set. This identifying data set contains unique identifying data, such as spectral data, corresponding to each compound present within the overlapped region. Errors in the identifying data set are minimized through an iterative process that modifies the estimated data set with each iteration. When the error is reduced to an acceptable level, the identifying data set serves to finally determine the number and identity of the compounds present within the region of overlap, i.e., resolve the overlapped peak, matrix techniques are advantageously used to handle and manipulate the large amount of data that is measured and generated. A computer may also advantageously be used to facilitate the speed and accuracy with which the matrices may be manipulated.
79 Citations
25 Claims
-
1. A method for resolving overlapped migration data obtained while performing qualitative analysis of a substance whose composition is to be determined, said method comprising the steps of:
-
(a) subjecting said substance to be analysed to a first apparatus adapted to separate compounds contained in said substance as a function of their respective migration time in a given medium, said first apparatus further adapted to provide a set of migration data indicating the migration behavior of said compounds as a function of time; (b) subjecting samples of said substance at specified times relative to the initiating of step (a) to a second apparatus, said second apparatus being adapted to measure and provide a unique set of characterizing time-invariant data for each of said samples, said set of characterizing data being a function of all the compounds contained in said sample at said specified times; (c) identifying regions of peak overlap within said set of migration data, said regions of overlap indicating the presence within said sample of at least two compounds having approximately the same differential migration times; (d) estimating a data set describing the differential migration time behavior of each of the compounds suspected of being present within each region of overlap identified in step (c), said set of estimated time behavior data describing the estimated shape of the migrational response of each said compounds within said region of overlap; (e) calculating a set of identifying time-invariant data for each of said compounds suspected of being present within said regions of overlap from said measured sets of characterizing data and said estimated sets of time behavior data; (f) varifying the goodness of the estimate made in step (d) by; (i) combining the calculated sets of identifying data derived in step (e) with the corresponding sets of estimated time behavior data made in step (d), (ii) comparing said combination of said estimated and calculated data to said measured set of characterizing data obtained in step (b), and (iii) determining an error factor based on said comparison; (g) modifying the estimated time behavior data made in step (d) as a function of the magnitude of the error factor determined in step (f); (h) repeating steps (e), (f), and (g) as many times as is necessary in order to reduce the error factor calculated in step (f)(iii) to an acceptable level; and (i) identifying the compounds present within said region of overlap based upon the last set of calculated identifying data obtained in step (e) as a result of the repetition of step (h). - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method for resolving overlapped chromatographic peaks of a substance being analysed, said substance having a plurality of unknown compounds mixed therewithin, said method comprising the steps of:
-
(a) injecting said substance to be analysed into a first end of a chromatographic column; (b) monitoring the output of said chromatographic column at a second end thereof with a first detector, said first detector being adapted to provide a first set of data indicating the migrational time behavior of the mixture of compounds through said column; (c) subjecting, in periodic fashion, samples of said substance to a second detector after said samples have passed through said column, said second detector being adapted to provide unique sets of time-invariant characterizing data, each set thereof being a function of the particular combination of said plurality of compounds preset within said sample at the time said sample is analysed by said second detector; (d) monitoring and recording said sets of time-invariant data and said first set of data using multi-channel detection means; (e) identifying regions of overlap within said first set of data wherein at least two of said plurality of compounds have migrated through the column within approximately the same time period; (f) assembling the sets of time-invariant data corresponding to said identified regions of overlap into a [D] matrix; (g) constructing a [C] matrix containing an estimate of the time behavior of each of said plurality of compounds within each identified region of overlap, said estimate being based upon time-behavior estimation means; (h) calculating an [A] matrix that is a function of said [D] and [C] matrices, said [A] matrix indicating a single set of time-invariant characterizing data for each of said plurality of compounds present within said region of overlap, the relationship between [A], [C], and [D] matrices being [A][C]=[D]; (i) testing the accuracy of said [A] matrix by multiplying said [A] and [C] matrices to form a [D'"'"'] matrix, and by comparing said [D'"'"'] matrix to said [D] matrix, the difference between the [D'"'"'] and [D] matrices indicating a measure of the error present in the [A] matrix; (j) miminizing the error in the [A] matrix by using said measure of the error determined in step (i) to guide a modification to the [C] matrix, using said modified [C] matrix to recalculate the [A] matrix as was done in step (h), and testing the accuracy of the re-calculated [A] matrix as was done in step (i); (k) repeating step (j) as many times as is required to reduce said error in the [A] matrix below a specified limit; and (1) using the [A] matrix having the error minimized therein as a identifier of the individual compounds of said plurality of compounds that are present within said substance within said region of overlap. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. A method for determining the number of compounds present within overlapped migration data of a substance having a plurality of unknown compounds mixed therewithin, said method comprising the steps of:
-
(a) subjecting said substance to first apparatus adapted to separate compounds contained in said substance as a function of their respective differential migration time in a given medium, said first apparatus further adapted to provide a set of migration data indicating the migration behavior of said compounds as a function of time; (b) subjecting samples of said substance at specified times relative to the initiating of step (a) to a second apparatus, said second apparatus being adapted to measure and provide a unique set of characterizing time-invariant data for each of said samples, said set of characterizing data being a function of all the compounds contained in said sample at said specified times; (c) identifying regions of peak overlap within said set of migration data, said regions of overlap indicating the presence within said sample of two or more compounds having approximately the same migration times; assuming the number of compounds present within each region of overlap identified in step (c) to be a low integer, n; (e) estimating a data set describing the differential migration time behavior of each of the compounds suspected of being present within each region of overlap identified in step (d), said set of estimated time behavior data describing the estimated shape of the migrational response of each said compounds within said region of overlap; (f) calculating a set of identifying time-invariant data for each of said compounds suspected of being present within said regions of overlap from said measured sets of characterizing data and said estimated sets of time behavior data; (g) determining an error factor representing the correctness of the number of compounds present within each region of overlap by; (i) combining the calculated sets of identifying data derived in step (f) with the corresponding sets of estimated time behavior data made in step (e), (ii) comparing said combination of said estimated and calculated data to said measured set of characterizing data obtained in step (b), and (iii) determining an error factor based on said comparison; (i) comparing the last determined error factor from the most recent performance of step (g) with the next-to-last determined error factor obtained from the next-to-most recent performance of step (g); (j) repeating steps (e), (f), (g), (h), and (i) until the error factors compared in step (i) remain essentially the same; and (k) assigning as the correct number of compounds present within the substance the last value of n from step (h).
-
-
22. An apparatus for resolving overlapped chromatographic peaks of a substance being analysed, said substance having a plurality of unknown compounds mixed therewithin that need to be identified, said apparatus comprising:
-
a chromatographic column; means for injecting said substance to be analysed into a first end of said column; first detector means coupled to a second end of said column for providing a first set of data indicating the migrational time behavior of said mixture of compounds through said column; second detector means coupled through an interface unit to said first detector means for providing a plurality of second sets of data, said second detector means adapted to analyse, in periodic fashion, samples of said substance after said samples have passed through said column and said interface unit, said analysis of said second detector means resulting in said second sets of data, each set thereof providing unique time-invariant characterizing data that is a function of the particular combination of said plurality of compounds present within said sample at the time said sample is analysed by said second detector means; multi-channel detection means for monitoring and recording said first set of data and said second sets of data; means for identifying regions of overlap within said first set of data wherein at least two of said plurality of compounds have migrated through the column within approximately the same time period; means for assembling the sets of time-invariant data recorded by said multi-channel detection means corresponding to said identified regions of overlap into a [D] matrix; time-behavior estimation means for constructing a [C] matrix containing an estimate of the time behavior of each of said plurality of compounds within each identified region of overlap; calculation means for calculating an [A] matrix that is a function of said [D] and [C] matrices, said [A] matrix indicating a single set of time-invariant identifying data for each of said plurality of compounds present within said region of overlap; error minimization means for minimizing any error contained within said [A] matrix as a result of inaccuries in the time behavior estimates of said [C] matrix, said means for minimizing error comprising; means for multiplying said [A] and [C] matrices to form a [D] product matrix, means for comparing the difference between said [D'"'"'] and [D] matrices, said difference indicating a measure of the error present within said [A] matrix, means for updating the timer behavior estimates contained in the [C] matrix as a function of the magnitude of said error found in said [A] matrix, and means for recalculating said [A] matrix based upon said updated [C] matrix; means for repeating the updating and recalculating the [C] and [A] matrices respectively using said error minimization means until the error contained within the [A] matrix is reduced below an acceptable level; and means for applying analysing the identifying data contained within the error-minimized [A] matrix to identify the specific compounds of said plurality of compounds that are present within said substance within said region of overlap. - View Dependent Claims (23, 24, 25)
-
Specification