NRfamPred: A proteome-scale two level method for prediction of nuclear receptor proteins and their sub-families

Abstract

Nuclear receptor proteins (NRP) are transcription factor that regulate many vital cellular processes in animal cells. NRPs form a super-family of phylogenetically related proteins and divided into different sub-families on the basis of ligand characteristics and their functions. In the post-genomic era, when new proteins are being added to the database in a high-throughput mode, it becomes imperative to identify new NRPs using information from amino acid sequence alone. In this study we report a SVM based two level prediction systems, NRfamPred, using dipeptide composition of proteins as input. At the 1st level, NRfamPred screens whether the query protein is NRP or non-NRP; if the query protein belongs to NRP class, prediction moves to 2nd level and predicts the sub-family. Using leave-one-out cross-validation, we were able to achieve an overall accuracy of 97.88% at the 1st level and an overall accuracy of 98.11% at the 2nd level with dipeptide composition. Benchmarking on independent datasets showed that NRfamPred had comparable accuracy to other existing methods, developed on the same dataset. Our method predicted the existence of 76 NRPs in the human proteome, out of which 14 are novel NRPs. NRfamPred also predicted the sub-families of these 14 NRPs.

Prediction of endoplasmic reticulum resident proteins using fragmented amino acid composition and support vector machine

“…In prediction methods, three cross-validation approaches are most frequently used: independent dataset test, subsampling test (N-fold cross-validation) and jack-knife test or leave-one-out cross-validation (LOOCV). Among the three cross-validation approaches, LOOCV gives unique result for a given benchmark dataset and hence used in a number of prediction methods ( Kumar et al, 2014b ; Kumar et al, 2015 ; Kumari, Kumar & Kumar, 2014 ; Lin et al, 2011 ; Xiao et al, 2013 ; Xu et al, 2013 ). In the present study, we have used LOOCV approach for the evaluation purpose, in which all except one sequence of dataset is used as training set and remaining one sequence as test set.…”

“…To evaluate the performance of each trained SVM model, we used standard parameters regularly used in other prediction methods for prediction evaluation namely sensitivity, specificity, accuracy and Matthews Correlation Coefficient (MCC) ( Kumar, Gromiha & Raghava, 2008 ; Kumar et al, 2014a ; Kumar et al, 2014b ; Kumar et al, 2015 ; Panwar, Arora & Raghava, 2014 ) as formulated below: where TP represents true positive (proteins, which are actually ERRPs and were also predicted as ERRPs), TN represents true negative (proteins, which are actually non-ERRPs and also predicted as non-ERRPs), FP represents false positive (the number of non-ERRP predicted as ERRPs), FN represents false negative (number of proteins, which are actually ERRPs but predicted as non-ERRPs).…”

“…Dipeptide composition is a modified form of amino acid composition, which has been used in a number of protein classification methods like nuclear receptor family classification ( Bhasin & Raghava, 2004 ; Kumar et al, 2014b ), membrane protein prediction, ion channels prediction ( Lin & Ding, 2011 ), protein fold prediction ( Shamim, Anwaruddin & Nagarajaram, 2007 ), heat shock protein and its class prediction ( Kumar, Gromiha & Raghava, 2011 ; Kumar, Kumari & Kumar, 2016 ; Reczko & Bohr, 1994 ). Dipeptide composition calculates the number of all possible types of amino acid pairs in a protein sequence in a sliding window mode with step size 1 and window size 2.…”

Large-scale identification of human protein function using topological features of interaction network

“…The detailed explanations and definitions can be found in reference38. All of these features have been widely employed to identify various properties of protein3940.…”

Accurate prediction of nuclear receptors with conjoint triad feature

“…Subsequently, Xiao et al [ 16 ] constructed a predicting model based on physical-chemical matrix via a series of auto-covariance and cross-covariance transformations, and resulting predictor achieved higher accuracy rates of recognition on the same dataset [ 15 ]. Recently, a proteome-scale two level predicting method, named “NRfamPred”, was developed based on dipeptide composition [ 17 ].…”