JNets: Exploring networks by integrating annotation

BMC Bioinformatics

Table 1 HIV-1 interactions with approved drug target genes, by HIV-1 element.

HIV-1 element	Total interactions	Interactions with drug target genes	Corrected p
Env-gp120	856 (22%)	239 (29%)	5.85 × 10^-6
Integrase	102 (3%)	5 (1%)	7.15 × 10^-4
Rev	109 (3%)	7 (1%)	1.36 × 10^-3
Env-gp41	190 (5%)	58 (7%)	4.65 × 10^-3
Tat	1394 (35%)	249 (30%)	2.50 × 10^-2
Capsid	45 (1%)	16 (2%)	6.23 × 10^-2
Vif	77 (2%)	8 (1%)	6.57 × 10^-2
Nef	302 (8%)	78 (10%)	7.60 × 10^-2
Pol	1 (< 1%)	0 (0%)	0.238
Vpr	275 (7%)	47 (6%)	0.173
Env-gp160	213 (5%)	49 (6%)	0.432
Env-gp41	190 (5%)	58 (7%)	1.07 × 10^-2
Rev	109 (3%)	7 (1%)	4.50 × 10^-3
Integrase	102 (3%)	5 (1%)	2.61 × 10^-3
Matrix	95 (2%)	15 (2%)	0.399
Retropepsin	89 (2%)	15 (2%)	0.506
Vif	77 (2%)	8 (1%)	0.120
Gag	72 (2%)	15 (2%)	0.889
Reverse transcriptase	45 (1%)	8 (1%)	0.605
Vpu	27 (1%)	5 (1%)	0.667
Nucleocapsid	26 (1%)	4 (< 1%)	0.486
p6	18 (< 1%)	2 (< 1%)	0.386
p1	3 (< 1%)	0 (0%)	0.253

The proportion of interactions that HIV-1 has with the products of genes that also encode drug targets, compared to total interactions, grouped by HIV-1 element are shown. The statistical significance of the difference in expected and actual figures, indicated by p-values, were calculated in JNets using two-tailed Chi-square tests, corrected for multiple testing using the Benjamini and Hochberg method [15]. Using a p cutoff of 0.05, we found significantly more Env-gp120 and Env-gp41 interactions and significantly fewer Tat, Integrase and Rev interactions among drug target genes, than would be expected at random.

ISSN: 1471-2105