Method and system for quantifying and removing spatial-intensity trends in microarray data

• US 7,302,348 B2
• Filed: 06/02/2004
• Issued: 11/27/2007
• Est. Priority Date: 06/02/2004
• Status: Active Grant

• Current Assignee: Agilent Technologies, Inc.
• Original Assignee: Agilent Technologies, Inc.
• Inventors: Ghosh, Jayati, Peck, Bill J., Zhou, Xiangyang, Corson, John F., Leproust, Eric M., Wolber, Paul K., Collins, Patrick J., Troup, Charles David, Delenstarr, Glenda Choate
• Primary Examiner(s): Brusca; John S
• Application Number:
  US10/859,868
• Publication Number:
  US 20050273268A1
• Time in Patent Office: 1,273 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 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 method and system for quantifying and correcting spatial-intensity trends for each channel of a microarray data set having one or more channels. The method and system of one embodiment of the present invention selects a set of features from each channel of the microarray data set. Based on the selected set of features, a surface is used to determine the intensities for all features in each channel of the microarray data set. Spatial-intensity trends within the microarray data set are quantified, based on the surface to the intensities for each channel of the microarray data set. After the surface has been determined, the spatial-intensity trend can be removed from the microarray data set.

Claims

1. A method for quantifying and correcting spatial-intensity trends in a microarray data set having one or more channels, the method comprising:

selecting a set of features, based on intensity level, from each channel of the microarray data set;

determining surface intensities for each feature in the channel of the microarray data set from a surface calculated based on intensity levels of the selected set of features;

quantifying the spatial-intensity trends in each channel of the microarray data set based on the surface intensities; and

correcting feature intensities for each channel of the multi-channel, microarray data set based on the determined surface intensities; and

performing at least one of;

outputting a result of the method to a display;

outputting a result of the method to memory;

outputting a result of the method to a computer on a network;

or outputting a result to a human user in human-readable format.

2. The method of claim 1 wherein selecting the set of features further includes selecting lowest-signal-intensity features or highest-signal-intensity features.

3. The method of claim 1 wherein the selected features are negative-control features or positive-control features.

4. The method of claim 2 wherein said selecting lowest, signal-intensity features, or highest-signal-intensity features further includes employing an optional window filter.

5. The method of claim 4 wherein employing the optional moving-window filter further comprises:

varying the moving-window filter size;

incrementally moving the moving-window filter; and

selecting a fraction of the lowest-signal-intensity features or the highest-signal intensity features that are within the boundaries of the moving-window filter.

6. The method of claim 5 wherein employing an optional moving-window filter further includes symmetrically extending boundary features of the microarray data set.

7. The method of claim 1 wherein selecting the set of features further includes excluding features having non-uniform intensity distributions.

8. The method of claim 1 wherein determining the surface for each feature further includes using locally-weighted, least-squares regression.

9. The method of claim 8 wherein using locally-weighted, least-squares regression further includes determining a user-defined neighborhood of filtered features associated with each feature.

10. The method of claim 9 wherein each neighborhood within the set of features is substantially the same size.

11. The method of claim 1 wherein selecting the set of features further includes employing a visual display to select non-clustering features.

12. The method of claim 1 wherein measuring the spatial-intensity trend in each channel of the multi-channel, microarray data set further includes determining, for each surface, one or more of:

a spatial root mean square filtered feature minus fit;

a spatial root mean square fit;

a normalized area;

a spatial volume;

a mean of the selected feature signal intensities; and

a root mean square of the selected feature signal intensities.

13. The method of claim 1 wherein said correcting feature intensities includes subtracting from each feature intensity a corresponding surface intensity.

14. The method of claim 1 wherein correcting feature intensities for each channel of the data set further includes:

normalizing the surface to the highest intensity value; and

dividing each feature intensity of the data set by the corresponding normalized surface intensity.

15. The method of claim 1 wherein quantifying and correcting spatial-intensity trends further includes quantifying and correcting spatial-intensity trends in algebraically transformed features intensities.

16. The method of claim 1, further comprising performing at least one of:

storing a spatial-intensity trend data set, produced by the method of claim 1, in a computer readable medium; and

transferring the spatial intensity trend data set to an intercommunicating entity via electronic signals.

17. The method of claim 1, further comprising storing results produced by the method of claim 1 in a computer-readable medium.

18. A method comprising communicating to a remote location a spatial-intensity trend obtained by a method of claim 1.

19. A method comprising receiving data produced by using the method of claim 1.

20. A physical computer readable medium carrying software to execute the method of claim 1.

21. A program on physical computer readable media configured to, when executed, provide a computer user interface comprising:

a listing of each microarray data set stored in a microarray database;

an optional debug mode;

a scrollable display listing available filters;

a selection window for selecting optional moving-window filtering;

an input window for inputting a size of the optional moving-window filter;

an input window for selecting a fraction of selected filtered features;

an input window for selecting an increment of the moving window filter; and

a display window for displaying metric values for each microarray data set that allows a user to determine which surfaces are added to the database.

22. A system for determining spatial intensity trends in microarrays data, the system comprising:

a computer processor;

a communications medium by which one or more microarray data sets are received by the microarray-data processing system;

a program, stored in the one or more memory components and executed by the computer processor to perform a method that selects a set of features, based on intensity levels, from each channel of the microarray data set;

determines surface intensities for each feature in the channel of the microarray data set based on the selected set of features;

quantifies the spatial-intensity trends in each channel of the microarray data set based on the surface intensities;

corrects feature intensities for each channel of the multi-channel, microarray data set based on the determined surface intensities; and

performs at least one of;

outputting a result of the method to a display;

outputting a result of the method to memory;

outputting a result of the method to a computer on a network;

or outputting a result to a human user in human-readable format.