The expansion of the human gene networks is based on the transcriptomic dataset provided by the FANTOM project. FANTOM5 promoter-level expression atlas come from RNA sequencing using single molecule CAGE (Cap Analysis Gene Expression) sequencing of extracted from 1,816 samples of different human tissues and cell lines. It contains expression profiles of 201,802 gene transcripts, corresponding to transcription initiations isoforms, indicated as p1@, p2@ etc. Unknown genes, i.e. without an annotated HGNC symbol, were excluded, reducing the dataset to a collection of 87554 transcripts, associated with 18889 genes, it thus contains plenty of information on human gene transcriptional profiles in different biological contexts that can be exploited for data mining for different purposes. To our best knowledge ours is the first attempt to infer genome-scale regulatory information from FANTOM5 data.
E. Blanzieri et al., "A Computing System for Discovering Causal Relationships Among Human Genes to Improve Drug Repositioning," in IEEE Transactions on Emerging Topics in Computing, vol. 9, no. 4, pp. 1667-1682, 1 Oct.-Dec. 2021, doi: 10.1109/TETC.2020.3031024.