MMGAT: a graph attention network framework for ATAC-seq motifs finding

BMC Bioinformatics

Table 1 Mean and standard deviation of precision, recall, F1_score, ACC, AUC, and PRC scores of six models on 180 human ATAC-seq datasets

Models	scFAN	Factornet	MMGraph	MMGraph + GL1	MMGraph + GL2	MMGAT
Precision	0.793 ± 0.066	0.801 ± 0.068	0.907 ± 0.037	0.911 ± 0.036	0.910 ± 0.030	0.925 ± 0.020
Recall	0.748 ± 0.093	0.781 ± 0.078	0.898 ± 0.043	0.910 ± 0.037	0.902 ± 0.036	0.921 ± 0.022
F1_score	0.732 ± 0.117	0.775 ± 0.084	0.897 ± 0.044	0.909 ± 0.037	0.901 ± 0.037	0.920 ± 0.022
ACC	0.748 ± 0.093	0.781 ± 0.078	0.898 ± 0.043	0.909 ± 0.037	0.901 ± 0.036	0.921 ± 0.022
AUC	0.874 ± 0.094	0.876 ± 0.070	0.962 ± 0.023	0.964 ± 0.021	0.963 ± 0.021	0.970 ± 0.017
PRC	0.878 ± 0.088	0.875 ± 0.072	0.956 ± 0.027	0.957 ± 0.028	0.957 ± 0.027	0.965 ± 0.022

ISSN: 1471-2105