System, method and software arrangement utilizing a multi-strip procedure that can be applied to gene characterization using DNA-array data

• US 7,957,908 B2
• Filed: 11/16/2004
• Issued: 06/07/2011
• Est. Priority Date: 11/17/2003
• Status: Active Grant

• Current Assignee: New York University
• Original Assignee: New York University
• Inventors: Mishra, Bud, Lerman, Gilad, McQuown, Joseph
• Primary Examiner(s): NEGIN, RUSSELL SCOTT
• Application Number:
  US10/579,811
• Publication Number:
  US 20070100556A1
• Time in Patent Office: 2,394 Days
• Field of Search: None
• US Class Current: 702/19
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16B 25/10 Gene or protein expression profiling; Expression-ratio estimation or normalisation
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/10 Signal processing, e.g. from mass spectrometry [MS] or from PCR

Abstract

A system, method and software arrangement are provided that use a fast adaptive multiscale procedure to characterize a random set of points spanning a high dimensional Euclidean space, and concentrated around special lower dimensional subsets. The procedure can be adapted to analyze gene expression data from microarray experiments, and may be applied generally to existing datasets without regard to whether a particular model exists to otherwise describe the dataset. The procedure accordingly can be used for identifying and mathematically isolating stable sets of data points in a given dataset from those in the same dataset that deviate from a stable model under various conditions.

Claims

1. A process for determining statistically-outlying data points present in at least one dataset, comprising:

a) receiving the at least one dataset;

b) determining at least one interval associated with the dataset;

c) using a hardware processing arrangement which comprises a processor, determining a plurality of subintervals of the at least one interval by iteratively dividing at least one of the at least one interval more than once until at least one predetermined criteria is met; and

d) determining the statistically-outlying data points present in the at least one dataset, wherein each data point is associated with a particular subinterval of the subintervals, and wherein the determination is performed based on information related to the subintervals and as a function of the length of the particular subinterval of the subintervals associated with the particular data point.

2. The process of claim 1, wherein the at least one dataset comprises data associated with levels of gene expression obtained under at least two different conditions.

3. The process of claim 2, wherein the different conditions reflect an occurrence of at least one of a physiological process, a pathophysiological process, an oncogenic process, a mutational process, a pharmacologically-induced process, an immuno-precipitation-induced process, or a developmental process.

4. The method of claim 1, wherein the at least one data set comprises data associated with financial trends.

5. The process of claim 1, further comprising:

at least one of displaying or storing the statistically-outlying data points in a storage arrangement in at least one of a user-accessible format or a user-readable format.

6. The process of claim 1, wherein at least one of the at least one predetermined criteria is a function of a portion of the dataset being contained in one of the subintervals.

7. The process of claim 6, wherein the data set comprises a total number of points, and wherein at least one of the at least one criteria is a function of a predetermined ratio of the statistically-outlying points to the total number of points.

8. The process of claim 1, wherein the subintervals are determined based on a dyadic grid division procedure.

9. The process of claim 1, wherein each of the statistically-outlying data points is further determined as a function of a distance of each point to a principal axis of the dataset.

10. The process of claim 1, wherein each particular subinterval of the subintervals comprises a respective region having a height, wherein the height is determined as a function of the length of the subinterval, and wherein each of the statistically-outlying data points is located outside of the respective region of the particular subinterval with which the statistically-outlying data point is associated.

11. The process of claim 10, wherein at least one of the at least one predetermined criteria is based on the number of points located within each region being below a predetermined threshold.

12. The process of claim 10, wherein the at least one dataset comprises a total number of points, wherein each particular subinterval of the subintervals comprises a total number of points, and wherein at least one of the at least one criteria is based on a constraint that a ratio of a number of points located outside each region to the total number of points in the particular subinterval exceeds a ratio of the number of statistically-outlying points in the dataset to the total number of points in the at least one dataset.

13. The process of claim 10, wherein at least one of the at least one predetermined criteria is based on an average distance of the total number of points located in each of the regions to a principal axis of the at least one dataset.

14. A non-transitory storage medium which includes thereon a software arrangement to be executed by a hardware processing arrangement for determining statistically-outlying data points present in at least one dataset, the software arrangement comprising:

a) a first set of instructions operable to configure the processing arrangement to receive the at least one dataset;

b) a second set of instructions operable to configure the processing arrangement to determine at least one interval associated with the dataset;

c) a third set of instructions operable to configure the processing arrangement to determine a plurality of subintervals of the at least one interval by iteratively dividing at least one of the at least one interval more than once until at least one predetermined criteria is met; and

d) a fourth set of instructions operable to configure the processing arrangement to determine the statistically-outlying data points present in the at least one dataset, wherein each data point is associated with a particular subinterval of the subintervals, and wherein the determination is performed based on information related to the subintervals and as a function of the length of the particular subinterval of the subintervals associated with the particular data point.

15. The storage medium of claim 14, wherein the at least one dataset comprises data associated with levels of gene expression obtained under at least two different conditions.

16. The storage medium of claim 15, wherein the different conditions reflect the occurrence of at least one of a physiological process, a pathophysiological process, an oncogenic process, a mutational process, pharmacologically-induced process, an immuno-precipitation-induced process, or a developmental process.

17. The storage medium of claim 14, wherein the at least one data set comprises data associated with financial trends.

18. The storage medium of claim 14, wherein the software arrangement further comprises:

a further set of instructions operable to configure the processing arrangement to at least one of display or store the statistically-outlying data points in a storage arrangement in at least one of a user-accessible format or a user-readable format.

19. The storage medium of claim 14, wherein at least one of the at least one predetermined criteria is a function of a portion of the at least one dataset being contained in one of the subintervals.

20. The storage medium of claim 19, wherein the at least one dataset comprises a total number of points, and wherein at least one of the at least one predetermined criteria is a function of a predetermined ratio of the statistically-outlying points to the total number of points.

21. The storage medium of claim 14, wherein the subintervals are determined based on a dyadic grid division procedure.

22. The storage medium of claim 14, wherein each of the statistically-outlying data points is further determined as a function of a distance of each point to a principal axis of the at least one dataset.

23. The storage medium of claim 14, wherein each particular subinterval of the subintervals comprises a region having a height, wherein the height is determined as a function of the length of the subinterval, and wherein each of the statistically-outlying data points is located outside of the region of the particular subinterval with which the statistically-outlying data point is associated.

24. The storage medium of claim 23, wherein at least one of the at least one predetermined criteria is based on the number of points located within each of the regions being below a predetermined threshold.

25. The storage medium of claim 23, wherein the data set comprises a total number of points, wherein each particular subinterval of the subintervals comprises a total number of points, and wherein at least one of the criteria is based on a constraint that a ratio of the number of points located outside each region to a total number of points in the particular subinterval exceeds a ratio of the number of statistically-outlying points in the at least one dataset to the total number of points in the at least one dataset.

26. The storage medium of claim 23, wherein at least one of the at least one predetermined criteria is based on an average distance of a total number of points located in each of the regions to a principal axis of the at least one dataset.

27. A system for determining statistically-outlying data points present in at least one dataset, comprising:

a hardware processing arrangement which includes a processor and which is operably configured to;

a) receive the at least one dataset;

b) determine at least one interval associated with the dataset;

c) determine a plurality of subintervals of the at least one interval by iteratively dividing at least one of the at least one interval more than once until at least one predetermined criteria is met; and

d) determine the statistically-outlying data points present in the at least one dataset, wherein each data point is associated with a particular subinterval of the subintervals, and wherein the determination is performed based on information related to the subintervals and as a function of the length of the particular subinterval of the subintervals associated with the particular data point.

28. The system of claim 27, further comprising a further hardware processing arrangement configured to generate the at least one dataset, wherein the further hardware processing arrangement comprises a further processor.

29. The system of claim 28, further comprising a detector configured to detect a plurality of signals indicative of gene expression and convert the detected signals into the at least one dataset.

30. The system of claim 27, wherein the processing arrangement is further operably configured to at least one of display or store the statistically-outlying data points in a storage arrangement in at least one of a user-accessible format or a user-readable format.

31. The system of claim 27, wherein the subintervals are determined based on a dyadic grid division procedure, wherein each particular subinterval of the subintervals comprises a region having a height, wherein the height is determined as a function of the length of the particular subinterval, wherein each of the statistically-outlying data points is located outside of the region of the particular subinterval with which the statistically-outlying data point is associated, wherein the at least one dataset comprises a total number of points, wherein each particular subinterval comprises a total number of points, and wherein one or more of the at least one predetermined criteria is based on at least one of (i) a portion of the at least one dataset being contained in one of the subintervals, (ii) a predetermined ratio of the statistically-outlying points to the total number of points, (iii) the number of points located within each of the regions being below a predetermined threshold, (iv) a constraint that a ratio of the number of points located outside each of the regions to the total number of points in the particular subinterval exceeds a ratio of the number of statistically-outlying points in the at least one dataset to the total number of points in the at least one dataset, or (v) an average distance of the total number of points located in each of the regions to a principal axis of the at least one dataset.

32. The system of claim 27, wherein the subintervals are determined based on a dyadic grid division procedure, wherein each particular subinterval of the subinterval comprises a region having a height, wherein the height is determined as a function of the length of the particular subinterval, wherein each of the statistically-outlying data points is located outside of the region of the particular subinterval to which the statistically-outlying data point is associated, wherein the at least one dataset comprises a total number of points, wherein each subinterval comprises a total number of points, and wherein one or more of the at least one predetermined criteria is based on at least two of (i) a portion of the at least one dataset being contained in one of the subintervals, (ii) a predetermined ratio of the statistically-outlying points to the total number of points, (iii) the number of points located within each of the regions being below a predetermined threshold, (iv) a constraint that a ratio of the number of points located outside each of the regions to the total number of points in the particular subinterval exceeds a ratio of the number of statistically-outlying points in the at least one dataset to the total number of points in the at least one dataset, or (v) an average distance of the total number of points located in each region to a principal axis of the at least one dataset.