Table 2.

Results of the KDVS for Parkinson disease study I & II

Kind	Enrichment		KDVS
Lists	PDvsN I	PDvsN II	PDvsN I	PDvsN II
Discr. GO terms	77	65	364	150
Discr. Genes	77	66	5705	4286
Comm. GO Terms	31	31	113	54
Comm. Genes	9	3	274	196
Benchmark Cov. GO Terms	2%	1%	25%	12%
Benchmark Cov. Genes	2%	1%	62%	44%

In the first two rows, the table shows the discriminant gene and GO term lists identified by enrichment–based approach and KDVS for Parkinson study I and II, respectively. Both studies were performed on two different microarray datasets of Parkinson (PD) and Normal (N) tissue samples. Discr. means discriminant.

In the next two rows, the table shows the gene and GO term lists obtained as an intersection between respective discriminant lists and benchmark lists, built for Parkinson’s disease. Comm. means common.

In the last two rows, the table shows the coverage of benchmark gene and GO term lists with respective discriminant lists. For KDVS, the coverage was calculated only for benchmark MF terms, according to the nature of classification task. Benchmark gene list consists of 444 elements. Benchmark GO term list consists of 2121 terms in total, including 446 MF terms. Bench. cov. means benchmark coverage.