Improving protein structural class prediction using novel combined sequence information and predicted secondary structural features

Abstract

Protein structural class prediction solely from protein sequences is a challenging problem in bioinformatics. Numerous efficient methods have been proposed for protein structural class prediction, but challenges remain. Using novel combined sequence information coupled with predicted secondary structural features (PSSF), we proposed a novel scheme to improve prediction of protein structural classes. Given an amino acid sequence, we first transformed it into a reduced amino acid sequence and calculated its word frequencies and word position features to combine novel sequence information. Then we added the PSSF to the combine sequence information to predict protein structural classes. The proposed method was tested on four benchmark datasets in low homology and achieved the overall prediction accuracies of 83.1%, 87.0%, 94.5%, and 85.2%, respectively. The comparison with existing methods demonstrates that the overall improvements range from 2.3% to 27.5%, which indicates that the proposed method is more efficient, especially for low-homology amino acid sequences.

Accurate prediction of protein structural classes by incorporating PSSS and PSSM into Chou's general PseAAC

“…We select the accuracy of each class and overall accuracy as evaluation indexes which are summarized in Table 6. The compared methods include the famous methods SCPRED [17] and MODAS [19], and the other competitive methods such as RKS-PPSC [18], IEA-PSSF [72], Zhang et al [45], PSSS-PSSM [20], LCC-PSSM [73], AADP-PSSM [21] and AAC-PSSM-AC [23]. Among these methods, the PSSS-PSSM method has the best overall accuracy on the three datasets.…”

An ensemble classifier based prediction of G-protein-coupled receptor classes in low homology

“…There are three cross-validation methods are often used to examine a predictor for its effectiveness in the statistical prediction: single test-set analysis, sub-sampling test, and jackknife test [9,8]. Owing to the most objective [3] and effective [11] to yield a unique result among the three cross-validation methods, elucidated by the recent studies [13,18,42], the jackknife test has been increasingly used on examining the accuracy of various predict methods.…”

“…In the present study, we depict 11 widely-used features S ¼ ðS 1 ; S 2 ; …; S 11 Þ T , which are designed to reflect the general contents and spatial arrangements of the secondary structural elements of a given protein sequence [11]. The 11-dimensional vector can be expressed as follows:…”

“…In the present study, we utilize the predicted secondary structural sequence (PSSF) [11] for representing the general contents and spatial arrangements of the secondary structural elements of a given GPCR sequence, which are more reliable than the exploration only based on the protein primary structure. We describe a novel PseAA composition with PSSF and approximate entropy (ApEn) of protein sequence to predict the classification of GPCRs in low homology.…”

Prediction of protein structural classes for low-similarity sequences using reduced PSSM and position-based secondary structural features

“…Since protein structural class concept was proposed by Levitt and Chothia (Levitt and Chothia, 1976;Andreeva et al, 2004;Murzin et al, 1995), various significant efforts have been made to predict protein structural class during the past 30 years (Dai and Wang, 2008;Chen et al, 2006;Chou, 2000;Kedarisetti et al, 2006;Dai et al, 2011). Previous studies indicated that protein structural classes could be predicted from amino acid sequences (Klein and Delisi, 1986;Chou, 1999;Chou and Shen, 2007), consequently, several features of protein sequences have been proposed for protein structural class prediction, such as short polypeptide composition (Luo et al, 2002;Sun and Huang, 2006;Zhang et al, 2014), pseudo AA composition (Ding et al, 2007;Wu et al, 2011;Liao et al, 2012;Kong et al, 2014) and collocation of function domain composition (Chou and Cai, 2004).…”