iProL: identifying DNA promoters from sequence information based on Longformer pre-trained model

BMC Bioinformatics

Table 1 Details of the E. coli promoter dataset

Dataset	Activity	Promoter							Non promoter	All
Dataset	Activity	\({\sigma }^{24}\)	\({\sigma }^{28}\)	\({\sigma }^{32}\)	\({\sigma }^{38}\)	\({\sigma }^{54}\)	\({\sigma }^{70}\)	\({\sigma }^{unknown}\)	Non promoter	All
Benchmark dataset	Strong	68	10	61	116	17	758	561	/	1591
	Weak	418	123	222	41	74	886	27	/	1791
		486	133	283	157	91	1644	588	/
		3382							3382
Independent test dataset	Strong	4	0	12	74	5	69	47	/	211
	Weak	30	5	17	11	0	98	8	/	169
	Confirmed	1	0	0	9	0	1	4	/	15
		35	5	29	94	5	168	59	/
		395							0

ISSN: 1471-2105