Fig. 10
From: Prior knowledge driven Granger causality analysis on gene regulatory network discovery
Prior knowledge “TF binding score” and “YeastNet” were separately tested on the YMC dataset using CGC-2SPR. The result regression coefficients (B ij ) are sampled from either the filtered gene set (average pairs) or the golden standard. Then the normalized histograms are plotted in this figure to compare the performance of CGC-2SPR with different prior knowledge. “TF binding score” prior knowledge affects golden standard distribution significantly with minor effects on average pair distribution. Respectively “YeastNet” only has barely recognizable effect on the golden standard distribution. a CGC-2SPR using “YeastNet”. b CGC-2SPR using “TF binding score”