Skip to main content

Table 1 Performance of prediction scores in the DREAM dataset

From: A simple gene set-based method accurately predicts the synergy of drug pairs

Scoring methods

PC-index

P-value

Co-gene score

\( \frac{g_{d_1,{d}_2}}{G} \)

0.648

<0.0001

Co-GS score

\( \frac{n_{d_{1,}{d}_2}}{L} \)

0.589

0.0036

Co-gene/GS score

\( \frac{{\displaystyle {\sum}_l}\frac{N_{d_1,{d}_2}^l}{N^l}}{n_{d_{1,}{d}_2}} \)

0.663

<0.0001

  1. Notations: \( {g}_{d_1,{d}_2} \), number of commonly regulated genes between drugs \( {d}_1 \) and \( {d}_2 \); \( G \), total number of genes; \( {n}_{d_{1,}{d}_2} \), number of gene sets with significant enrichment in both drugs; \( L \), total number of gene sets; \( {N}_{d_1,{d}_2}^l \), number of genes with significantly changes in both drug treatments within a co-enriched gene set l