System and method for biological data analysis using a bayesian network combined with a support vector machine

1. A method for analyzing biological microarray data and protein mass spectra data comprising the steps of:

classifying said microarray data in a first classifier;

classifying said protein mass spectra data in a second classifier;

combining the results of the first classifier with the results of the second classifier; and

analyzing the results as a function of the similarity measure of the first classifier and the similarity measure of the second classifier.

2. The method of claim 1 , wherein the first set of biological data and the second set of biological data are the same.

3. The method of claim 1 , wherein the first classifier is a support vector representation and discrimination machine.

4. The method of claim 1 , wherein the second classifier is a Bayesian network.

5. The method of claim 4 , wherein said Bayesian network comprises computing mutual information of pairs of data of said data set, creating a draft network based on the mutual information, wherein data item of said data set comprise nodes of said network and the edges connecting a pair of data nodes represent the mutual information of said nodes, thickening said network by adding edges when pairs of data nodes cannot be d-separated, and thinning said network by analyzing each edge of said draft network with a conditional independent test and removing said edge if said corresponding data nodes can be d-separated.

6. The method of claim 1 , wherein the results of the first classifier and the second classifier are combined in parallel.

7. The method of claim 1 , wherein said combining step comprises weighing the results of the first and second classifiers based on the input patterns.

8. A program storage device readable by a computer, tangibly embodying a program of instructions executable by the computer to perform the method steps for analyzing biological microarray data and protein mass spectra data, said method comprising the steps of:

classifying said microarray data in a first classifier;

classifying said protein mass spectra data in a second classifier;

combining the results of the first classifier with the results of the second classifier, and

analyzing the results as a function of the similarity measure of the first classifier and the similarity measure of the second classifier.

9. The computer readable program storage device of claim 8 , wherein the first set of biological data and the second set of biological data are the same.

10. The computer readable program storage device of claim 8 , wherein the first classifier is a support vector representation and discrimination machine.

11. The computer readable program storage device of claim 8 , wherein the second classifier is a Bayesian network.

12. The computer readable program storage device of claim 11 , wherein said Bayesian network comprises computing mutual information of pairs of data of said data set, creating a draft network based on the mutual information, wherein data item of said data set comprise nodes of said network and the edges connecting a pair of data nodes represent the mutual information of said nodes, thickening said network by adding edges when pairs of data nodes cannot be d-separated, and thinning said network by analyzing each edge of said draft network with a conditional independent test and removing said edge if said corresponding data nodes can be d-separated.

13. The computer readable program storage device of claim 8 , wherein the results of the first classifier and the second classifier are combined in parallel.

14. The computer readable program storage device of claim 8 , wherein said combining step comprises weighing the results of the first and second classifiers based on the input patterns.

15. A method for analyzing biological data comprising the steps of:

classifying a set of microarray data in a first classifier, wherein said first classifier is a support vector representation and discrimination machine, wherein said machine discriminates said data into a plurality of classes using a plurality of discrimination functions, wherein an inner product of each said discrimination function with a kernel function is evaluated on said data, wherein the norm of each said discrimination is minimized, and wherein the value of each said inner product is compared to a threshold to determine whether a microarray data item belongs to a class associated with said discrimination function;

classifying a set of protein mass spectra data in a second classifier, wherein said second classifier is a Bayesian network; and

analyzing the combined results of said first classifier and said second classifier as a function of the similarity measure of the first classifier and the similarity measure of the second classifier.