Functional discrimination of membrane proteins using machine learning techniques

BMC Bioinformatics

Table 4 Discrimination of channels/pores, electrochemical potential-driven transporters and primary active transporters using different machine learning approaches with amino acid occurrence as features

Method	5-fold cross-validation
	Sensitivity			Precision			F-Measure			Accuracy
	F1	F2	F3	F1	F2	F3	F1	F2	F3	(%)
Bayesnet	0.329	0.735	0.567	0.554	0.515	0.572	0.413	0.606	0.569	54.6
Naive Bayes	0.202	0.757	0.575	0.477	0.512	0.534	0.284	0.611	0.554	51.8
Logistic function	0.533	0.713	0.705	0.689	0.717	0.604	0.601	0.715	0.651	65.
RBF network	0.247	0.727	0.633	0.486	0.593	0.530	0.328	0.654	0.577	54.6
Support vector machine	0.163	0.727	0.826	0.847	0.705	0.529	0.273	0.716	0.645	60.0
k-nearest neighbor	0.629	0.705	0.640	0.634	0.683	0.651	0.632	0.694	0.646	65.6
Bagging meta learning	0.553	0.685	0.737	0.676	0.733	0.625	0.608	0.709	0.676	66.7
Classification via Regression	0.465	0.721	0.721	0.686	0.702	0.602	0.547	0.711	0.656	64.5
Decision tree J4.8	0.543	0.625	0.555	0.526	0.592	0.593	0.534	0.609	0.574	57.2
NBTree	0.471	0.570	0.659	0.553	0.656	0.548	0.508	0.610	0.598	57.7
Partial decision tree	0.520	0.647	0.623	0.551	0.645	0.600	0.535	0.646	0.612	60.0
Neural network	0.549	0.701	0.761	0.695	0.780	0.622	0.613	0.739	0.684	68.1
Jack-knife test	0.500	0.703	0.729	0.639	0.749	0.607	0.561	0.726	0.663	65.4
Equal data	0.723	0.743	0.574	0.691	0.712	0.630	0.707	0.727	0.601	68.0

F1: channels/pores; F2: electrochemical potential-driven transporters; F3: primary active transporters. Equal data: Results obtained with a dataset of 502 proteins each in all the three classes of transporters.

ISSN: 1471-2105