Journal: Bioinformatics Advances
Article Title: TransPilot: mining key transcription factors by correlating binding sites with differentially expressed genes
doi: 10.1093/bioadv/vbag145
Figure Lengend Snippet: The schema of TransPilot server. (A) LLR-matrix coding for positive training nucleotide sequences. (B) LLR-matrix coding for negative training nucleotide sequences. (C) The ANN model and creation of the ranked and directed TF-target dataset. (D) Creation of ranked gene list from transcriptome data, identification of key TFs by TransPilot, and finally build the regulation network.
Article Snippet: The TF-target pairs were ranked by the ANN score ( Y ANN , > 0.5) and the directions (positive or negative) of the regulatory relationships for each pair were determined by the correlation coefficient between their expression levels in a reference transcriptome dataset obtained from the Human Protein Atlas database ( Uhlén et al. 2015 ).
Techniques: