GPRED-GC: a Gene PREDiction model accounting for 5 ′- 3′ GC gradient

BMC Bioinformatics

Table 7 The comparison of the corresponding parameters in the two HMMs for these two sets of cutoffs

From	To	lowT=0.30, highT=0.50		lowT=0.31, highT=0.52
		Transition probabilities	Training count	Transition probabilities	Training count
ASS⁰	\(E^{1}_{\mathrm {H}}\)	0.051961	212	0.045098	184
ASS⁰	\(E^{1}_{\mathrm {M}}\)	0.125490	510	0.132353	540
ASS⁰	\(E^{1}_{\mathrm {L}}\)	0.000980	4	0.000980	4
ASS⁰	\(E^{2}_{\mathrm {H}}\)	0.034314	140	0.027450	112
ASS⁰	\(E^{2}_{\mathrm {M}}\)	0.124510	508	0.129412	528
ASS⁰	\(E^{2}_{\mathrm {L}}\)	0.000980	4	0.002941	12
ASS¹	\(E^{0}_{\mathrm {H}}\)	0.119469	216	0.101770	184
ASS¹	\(E^{0}_{\mathrm {M}}\)	0.316372	572	0.334071	604
ASS¹	\(E^{0}_{\mathrm {L}}\)	0.004425	8	0.004425	8
ASS¹	\(E^{\mathrm {H}}_{\text {term}}\)	0.066372	120	0.055310	100
ASS¹	\(E^{\mathrm {M}}_{\text {term}}\)	0.130531	236	0.139381	252
ASS¹	\(E^{\mathrm {L}}_{\text {term}}\)	0.002212	4	0.004425	8
rDSS⁰	\(rE^{2}_{\mathrm {H}}\)	0.107246	148	0.092754	128
rDSS⁰	\(rE^{2}_{\mathrm {M}}\)	0.272464	376	0.284058	392
rDSS⁰	\(rE^{2}_{\mathrm {L}}\)	0	0	0.002899	4
rDSS¹	\(rE^{2}_{\mathrm {H}}\)	0.074074	96	0.067901	88
rDSS¹	\(rE^{2}_{\mathrm {M}}\)	0.379630	492	0.385802	500
rDSS¹	\(rE^{2}_{\mathrm {L}}\)	0.003086	4	0.003086	4
rDSS²	\(rE^{2}_{\mathrm {H}}\)	0.099088	348	0.077449	272
rDSS²	\(rE^{2}_{\mathrm {M}}\)	0.407745	1432	0.428246	1504
rDSS²	\(rE^{2}_{\mathrm {L}}\)	0.001139	4	0.002278	8

Set1: lowT and highT are 0.30 and 0.50. Set2: lowT and highT are 0.31 and 0.52. The different probabilities before using pseudocount and their corresponding training counts are listed

ISSN: 1471-2105