PEPred-Suite: improved and robust prediction of therapeutic peptides using adaptive feature representation learning

Abstract

Motivation Prediction of therapeutic peptides is critical for the discovery of novel and efficient peptide-based therapeutics. Computational methods, especially machine learning based methods, have been developed for addressing this need. However, most of existing methods are peptide-specific; currently, there is no generic predictor for multiple peptide types. Moreover, it is still challenging to extract informative feature representations from the perspective of primary sequences. Results In this study, we have developed PEPred-Suite, a bioinformatics tool for the generic prediction of therapeutic peptides. In PEPred-Suite, we introduce an adaptive feature representation strategy that can learn the most representative features for different peptide types. To be specific, we train diverse sequence-based feature descriptors, integrate the learnt class information into our features, and utilize a two-step feature optimization strategy based on the area under receiver operating characteristic curve to extract the most discriminative features. Using the learnt representative features, we trained eight random forest models for eight different types of functional peptides, respectively. Benchmarking results showed that as compared with existing predictors, PEPred-Suite achieves better and robust performance for different peptides. As far as we know, PEPred-Suite is currently the first tool that is capable of predicting so many peptide types simultaneously. In addition, our work demonstrates that the learnt features can reliably predict different peptides. Availability and implementation The user-friendly webserver implementing the proposed PEPred-Suite is freely accessible at http://server.malab.cn/PEPred-Suite. Supplementary information Supplementary data are available at Bioinformatics online.

Better understanding and prediction of antiviral peptides through primary and secondary structure feature importance

“…Further, based on the assumption that AVPs have low sequence similarity the use of pseudo amino acid composition (PseAAC) 26 was introduced as AVP peptide features in the AdaBoost machine learning model 23 . In recent years ensemble-based methods have been introduced, such as Meta-iAVP 25 and PePred-Suite 24 . The Meta-iAVP approach uses machine learning to transform the feature space into a modified 6-dimensional predicted output vector, which then becomes the input data to the meta-classifier to predict the class of validation data set.…”

iQSP: A Sequence-Based Tool for the Prediction and Analysis of Quorum Sensing Peptides Using Informative Physicochemical Properties

“…From Table 5, we can observe that QSPpred-FL yields the highest prediction performance of 94.30% Ac and 0.885 MCC over 10-fold CV, while our proposed model iQSP gave a 91.07 ± 1.77% Ac and 0.82 ± 0.04 MCC. On the other hand, based on the independent validation test, iQSP outperformed that other methods with 93.00 ± 1.97% Ac, 0.86 ± 0.04 MCC and 0.96 ± 0.02 auROC, which was better than the existing QSP predictors [28,29,48]. Although, iQSP achieved slightly better than QSPpred-FL, our proposed model showed significant improvement than QSPpred-FL considering the two objectives: using the less complexity of prediction methods (1 SVM vs. 99 RFs) and a minimum number of features used (18D vs. 913D).…”

“…To demonstrate the effectiveness and power of our method, we conducted a comparative study of our final model (named iQSP) with the existing methods. To date, there are only two existings methods developed for the prediction of QSPs, i.e., QSPpred [29] and QSPpred-FL [28,48], performing on the benchmark and independent datasets over 10-fold CV and independent validation test. Table 5 lists the preformance comparisons of iQSP and the existing methods.…”

“…To make a fair comparison with the existing methods, a series of comparative experiments were carried out, while the same benchmark (200 QSPs and 200 Non-QSPs) and independent (20 QSPs and 20 Non-QSPs) datasets were used to develop and investigate the efficiency and effectiveness of our proposed QSP predictor. However, previous studies [28,29,48] performed these two datasets using a single random sampling procedure. As elaborated in [39,40,42,43,45,49,50], this procedure might find a possible bias of the random sampling process and provided a good predictive result by chance.…”

Identification of Secreted Proteins From Malaria Protozoa With Few Features

“…The ''classification'' option of Weka offers a variety of classifier patterns, such as randomforest, zeroR, kstar and libsvm. Random Forest has been widely utilized in bioinformatics [45]- [52]. In this paper, random forest (rf) was adopted as a classifier, and 10-fold cross-validation is used to determine its performance.…”