id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_pz3pacjpbrf6hm67aqomk5ag6e	Ghazaleh Khodabandelou	Genome annotation across species using deep convolutional neural networks	2020	18	.pdf	application/pdf	7826	754	61	We here address this issue by assessing the genomewide performance of networks trained with sets exhibiting different ratios of positive be used to predict gene-start sites in a related species, when using training sets providing Keywords Transcription start sites, Promoters, Genome annotation, Deep learning, DNA motifs, A CNN (see Fig. 1) is trained in order to predict the presence of a GSS in a DNA sequence Training models for genome annotation of GSS model applied to human chromosome X still captures most of the GSS and gives a λ score (C) Lambda values computed for networks trained on each species non-X chromosome GSS (t) Lambda scores are computed from predictions done on GSS of (A) human, (B) mouse, (C) model trained on the chicken genome performs less well when applied on the mammalian use the X chromosomes of human and mouse GSS as a case study, and apply models	./cache/work_pz3pacjpbrf6hm67aqomk5ag6e.pdf	./txt/work_pz3pacjpbrf6hm67aqomk5ag6e.txt