MISCORE: a new scoring function for characterizing DNA regulatory motifs in promoter sequences

Wang, Dianhui; Tapan, Sarwar

doi:10.1186/1752-0509-6-S2-S4

BMC Systems Biology

Table 7 Recognizability scores for the best candidate motifs using pk models

From: MISCORE: a new scoring function for characterizing DNA regulatory motifs in promoter sequences

Result details: a 10-run average μ score on each dataset
data group (dg)	TF	R _pk	$R_{l_{p k}}$	PCC	ALLR	KLD	ED	SW
	CREB	0.339	0.333	0.096	0.295	0.275	0.370	0.080
	SRF	0.667	0.717	0.500	0.553	0.553	0.657	0.564
	TBP	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	MEF2	1.000	1.000	1.000	1.000	1.000	1.000	1.000
dg ₁	MYOD	0.645	0.651	0.665	0.656	0.656	0.656	0.640
	ERE	1.000	1.000	1.000	1.000	0.917	0.875	1.000
	E2F	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	CRP	1.000	1.000	1.000	1.000	1.000	1.000	0.792
	avg	0.831	0.837	0.783	0.813	0.800	0.820	0.760
	dm01g	0.667	0.667	0.342	0.528	0.694	0.722	0.371
	dm04m	0.377	0.485	0.662	0.498	0.487	0.484	0.647
	hm02r	0.800	0.700	1.000	0.547	0.447	0.447	1.000
	hm03r	0.255	0.425	0.690	0.514	0.514	0.300	0.556
	hm06g	0.444	0.429	0.611	0.407	0.353	0.546	0.427
	hm08m	0.861	0.861	0.852	0.854	0.771	0.857	0.857
	hm09g	0.539	0.565	0.205	0.389	0.512	0.556	0.285
	hm10m	0.412	0.495	0.558	0.490	0.490	0.500	0.820
dg ₂	hm11g	0.302	0.329	0.829	0.335	0.285	0.333	0.829
	hm16g	0.690	0.767	0.105	0.617	0.767	0.900	0.100
	hm17g	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	hm20r	0.537	0.537	0.708	0.542	0.542	0.548	0.708
	hm21g	0.148	0.148	0.483	0.204	0.214	0.214	0.324
	hm24m	0.573	0.650	1.000	0.592	0.592	0.725	0.867
	hm26m	0.450	0.650	0.369	0.650	0.567	0.617	0.700
	mus02r	0.182	0.209	0.329	0.184	0.184	0.199	0.345
	mus10g	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	mus11m	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	yst08r	0.567	0.633	0.524	0.567	0.583	0.580	0.767
	yst09g	0.201	0.232	0.292	0.179	0.186	0.217	0.321
	avg	0.550	0.589	0.628	0.555	0.559	0.587	0.646
	CREB	0.642	0.642	0.556	0.657	0.657	0.667	0.476
	SRF	0.667	0.667	0.523	0.707	0.650	0.667	0.822
dg ₃	TBP	1.000	1.000	1.000	1.000	1.000	1.000	1.000
	MEF2	0.653	0.656	0.656	0.750	0.850	0.662	0.482
	MYOD	0.486	0.653	0.500	0.563	0.563	0.577	0.661
	avg	0.690	0.723	0.647	0.735	0.744	0.715	0.688
Result summary: a 10-run average μ score on each data group
dg ₁		0.831	0.837	0.783	0.813	0.800	0.820	0.760
dg ₂		0.550	0.589	0.628	0.555	0.559	0.587	0.646
dg ₃		0.690	0.723	0.647	0.735	0.744	0.715	0.688
avg		0.690	0.717	0.686	0.701	0.701	0.707	0.698

Remark: MISCORE metrics R_pk and $R_{l_{p k}}$ compute motif-to-pk similarity through the characterization of the motif signals, while the other metrics can not perform motif characterization. The result summary shows that MISCORE is capable of effectively utilizing the pk models in recognizing the functional motifs. Note: PCC: Pearson correlation coefficient [42]; ALLR: average log likelihood ratio [41]; KLD: Kullback-Leibler divergence [43–45]; ED: Euclidean distance [46]; and SW: Sandeline-Wasserman metric [47].

Back to article page

ISSN: 1752-0509

Contact us

General enquiries: ORSupport@springernature.com