System and methods for non-targeted processing of chromatographic data

A method for processing hyphenated chromatographic data to delineate components of a sample, the method comprising:

• a) obtaining hyphenated chromatographic data points for a sample that each comprise at least three dimensions, one of the dimensions being a response dimension and one of the dimensions being a continuous dimension;
• b) subjecting at least a portion of the data points to an algorithm that:
• i) partitions the data points into a high response dimension group and a low response dimension group according to a designated response dimension value threshold;
• ii) organizes the data points of the high response dimension group into discrete clusters according to the high response dimension group data points' continuous dimension values by starting at either the smallest or largest continuous dimension value associated with data points that are not yet associated with a discrete cluster and delineating, at the largest gap between continuous dimension values of adjacent data points within a predetermined resolution window, a boundary between discrete clusters, wherein the clusters of high response dimension group data points are of varying width of less than or equal to the width of the resolution window and wherein at least some of the clusters of high response dimension group data points are indicative of components of the sample, wherein step b) is repeated until all data points of the high response dimension group are associated with a cluster;
• and
• c) outputting to a user information describing the clusters in a user-readable format, wherein the steps a), b), and c) are performed on a suitably-configured computer and wherein the algorithm functions according to the following steps:
• i. creating an ordered set of the continuous dimension values of the data points within the high response dimension group;
• ii. selecting, as a start value, either the smallest or the largest value within the ordered set;
• iii. finding, as an end value, a value within the ordered set that is the closest to the start value from among the values having a distance from the start value that is greater than a designated resolution limit;
• iv. designating as a subset the values of the ordered set comprising the start value, the end value, and every value lying between;
• v. locating a largest gap between any consecutive two of the values of the subset, and designating the two consecutive values as a stop value and restart value, the stop value being the closer of the two to the start value;
• vi. marking as a cluster the high response dimension group data points having a continuous dimension value ranging from the start value up to and including the stop value;
• and
• vii. repeating steps (iii-vi) using the restart value as the start value, until each data point is contained within a marked cluster.

A method for processing hyphenated chromatographic data to delineate components of a sample, the method comprising:

• a) obtaining hyphenated chromatographic data points for a sample that each comprise at least three dimensions, one of the dimensions being a continuous dimension and one of the dimensions being a discrete dimension;
• b) subjecting at least a portion of the data points to an algorithm that:
• i) organizes the data points within each discrete dimension value into clusters according to the data points' continuous dimension values by starting at either the smallest or largest continuous dimension value associated with data points that are not yet associated with a discrete cluster and delineating, at a largest gap between continuous dimension values of adjacent data points within a predetermined resolution window, a boundary between discrete clusters, wherein the clusters of data points are of varying width of less than or equal to the width of the resolution window, wherein step i) is repeated until all data points are associated with a cluster;
• ii) for each cluster of data points, calculates a single representative continuous dimension value for the cluster and associates the calculated representative continuous dimension value to each data point in the cluster;
• and
• iii) organizes the data points, regardless of discrete dimension value, into groups according to their representative continuous dimension values, wherein the groups are of varying width of less than or equal to the width of the resolution window and wherein at least some of the groups are indicative of components of the sample;
• and
• c) outputting to a user information describing the groups in a user-readable format, wherein the steps a), b), and c) are performed on a suitably-configured computer and wherein the algorithm functions according to the following steps:
• i. creating an ordered set of the continuous dimension values of data points having a first discrete dimension value;
• ii. selecting, as a start value, either the smallest or the largest value within the ordered set and finds, as an end value, a value within the ordered set that is the closest to the start value from among the values having a distance from the start value that is greater than a designated resolution limit;
• iii. designating as a subset the values of the ordered set comprising the start value, the end value, and every value lying between;
• iv. locating a largest gap between any consecutive two of the values of the subset, and designating the two consecutive values as a stop value and a restart value, the stop value being the closer of the two to the start value;
• v. marking as a cluster the data points having a continuous dimension value ranging from the start value up to and including the stop value;
• vi. repeating steps (ii-v) using the restart value as the start value, until each data point having the first discrete dimension value is contained within a marked cluster;
• vii. repeating steps (i-vi) for each of the discrete dimension values of the remaining original data points;
• viii. calculating for each cluster of data points of step (v) a single representative continuous dimension value, associating the calculated representative continuous dimension value to each data point in the cluster, and ordering by value the calculated single representative continuous dimension values;
• and
• ix. repeating steps (ii-vi) using the single representative continuous dimension values until the number of the remaining representative values is zero, resulting in data points organized into groups according to consecutive clusters of the representative values, the clusters having varying width of less than or equal to the width of the resolution window.

A method for processing hyphenated chromatographic data to delineate components of a sample, the method comprising:

• a) obtaining hyphenated chromatographic data points for a sample that each comprise a continuous dimension, a discrete dimension, and a response dimension;
• b) subjecting at least a portion of the data points to an algorithm that:
• i) partitions the data points into a high response dimension group and a low response dimension group according to a designated response dimension value threshold;
• ii) organizes the high response dimension group of data points having a first discrete dimension value into discrete clusters according to the high response dimension group data points' continuous dimension values by starting at either the smallest or largest continuous dimension value associated with data points that are not yet associated with a discrete cluster and delineating at the largest gap between continuous dimension values of adjacent data points within a predetermined resolution window, a boundary between discrete clusters, wherein the clusters of high response dimension group data points are of varying width of less than or equal to the width of the resolution window, wherein step i) is repeated until all data points are associated with a cluster;
• iii) for each cluster of data points, calculates a single representative continuous dimension value for the cluster and associates the calculated representative continuous dimension value to each data point in the cluster;
• iv) repeats steps (ii) and (iii) over the entire range of the discrete dimension values;
• and
• v) organizes the data points, regardless of discrete dimension value, into groups according to their representative continuous dimension values wherein the groups are of varying width of less than or equal to the width of the resolution window and wherein at least some of the clusters are indicative of biochemical components of the sample;
• and
• c) outputting to a user information describing the groups in a user-readable format, wherein the steps a), b), and c) are performed on a suitably-configured computer and wherein the algorithm functions according to the following steps:
• i. partitioning the data points into a high response dimension group and a low response dimension group according to a designated response dimension value threshold;
• ii. creating an ordered set of the continuous dimension values of data points in the high response dimension group having a first discrete dimension value;
• iii. selecting, as a start value, either the smallest or the largest value within the ordered set and finds, as an end value, a value within the ordered set that is the closest to the start value from among the values having a distance from the start value that is greater than a designated resolution limit;
• iv. designating as a subset the values of the ordered set comprising the start value, the end value, and every value lying between;
• v. locating a largest gap between any consecutive two of the values of the subset, and designating the consecutive values as a stop value and a restart value, the stop value being the closer of the two to the start value;
• vi. marking as a cluster the data points having a continuous dimension value ranging from the start value up to and including the stop value;
• vii. repeating steps (iii-vi) using the restart value as the start value, until each data point in the high response dimension group having the first discrete dimension value is contained within a marked cluster;
• viii. repeating steps (ii-vii) for each of the discrete dimension values of the data points remaining in the high response dimension group;
• ix. calculating for each cluster of data points of step (vi) a single representative continuous dimension value, associating the calculated representative continuous dimension value to each high response dimension group data point in the cluster, and ordering by value the single representative continuous dimension values;
• and
• x. repeating steps (iii-vii) using the single representative continuous dimension values until the number of the remaining representative values is zero, resulting in high response dimension group data points organized into groups according to consecutive clusters of the representative values, the clusters having a varying width of less than or equal to the width of the resolution window.