RETRACTED: Analysis and prediction of β-Turn types using multinomial logistic regression and artificial neural network

Abstract

Motivation: So far various statistical and machine learning tech-niques applied for prediction of-Turns. The majority of these tech-proteins. We developed a hybrid approach for analysis and predic-niques have been only focused on the prediction of-Turn location in tion of different types of-Turn.The f3-Turn types I, II, IV and VIII. Multinomial logistic Results: A two-stage hybrid model developed to predict regression was initially used for the first time to select sig-nificant parameters in prediction of 13-Turn types using a self-consistency test procedure. The extracted parameters were consisted of 80 amino acid positional occurrences and 20 amino acid percentages in-Tum sequence. The most sig-logistic regression model. Among these, the occurrences nificant parameters were then selected using multinon1gnition glutamine, histidine, glutamic acid and arginine, respec-tively, in positions i, i+l, i+2 and i+3 of 13-Turn seque e had an overall relationship with 5 13-Turn types. A neural n model was then constructed and fed by the par-predictor. The networks have been trained a d a lected by mtiltinomial logistic regression rid non-homologous dataset of 565 protein chain fold cross-validation. It has been observed that h id model gives a Matthews correlation coefficient (M f 0.473 and 0.124, respectively, for 13-Turn typ III which odd also are best among previously reported resulurm distinguished the different types embedded binary logit comparisons which heied out so far.