RPiRLS: Quantitative Predictions of RNA Interacting with Any Protein of Known Sequence

Abstract

RNA-protein interactions (RPIs) have critical roles in numerous fundamental biological processes, such as post-transcriptional gene regulation, viral assembly, cellular defence and protein synthesis. As the number of available RNA-protein binding experimental data has increased rapidly due to high-throughput sequencing methods, it is now possible to measure and understand RNA-protein interactions by computational methods. In this study, we integrate a sequence-based derived kernel with regularized least squares to perform prediction. The derived kernel exploits the contextual information around an amino acid or a nucleic acid as well as the repetitive conserved motif information. We propose a novel machine learning method, called RPiRLS to predict the interaction between any RNA and protein of known sequences. For the RPiRLS classifier, each protein sequence comprises up to 20 diverse amino acids but for the RPiRLS-7G classifier, each protein sequence is represented by using 7-letter reduced alphabets based on their physiochemical properties. We evaluated both methods on a number of benchmark data sets and compared their performances with two newly developed and state-of-the-art methods, RPI-Pred and IPMiner. On the non-redundant benchmark test sets extracted from the PRIDB, the RPiRLS method outperformed RPI-Pred and IPMiner in terms of accuracy, specificity and sensitivity. Further, RPiRLS achieved an accuracy of 92% on the prediction of lncRNA-protein interactions. The proposed method can also be extended to construct RNA-protein interaction networks. The RPiRLS web server is freely available at .

A p53/lnc‐Ip53 Negative Feedback Loop Regulates Tumor Growth and Chemoresistance

“…Moreover, the level of ubiquitylated HDAC1 was increased by silnc‐Ip53 (Figure 6C) but decreased by lnc‐Ip53 overexpression (Figure 6D). Both lncPro [ 14 ] and RPiRLS [ 15 ] algorithms predicted potential interaction between lnc‐Ip53 and HDAC1, which was then verified experimentally. RNA immunoprecipitation (RIP) assay disclosed that compared with Flag‐control group, the Flag‐HDAC1‐precipitated complex contained more lnc‐Ip53 but similar amount of negative controls, like 18S rRNA or U6 RNA (Figure 6E; Figure S6A, Supporting Information).…”

Recent methodology progress of deep learning for RNA–protein interaction prediction

“…CTF + CGR designs representations derived from chaos game (H. Wang & Wu, 2018). RPiRLS extracts complex features by taking the contextual information into account using kernels (Shen et al, 2018). Besides, since constructing negative samples has a huge impact on model performance, PRIPU proposes a one-class method using positive samples and unlabeled samples (Cheng et al, 2015).…”

ccPDB 2.0: an updated version of datasets created and compiled from Protein Data Bank

“…Since the first version of ccPDB (18) published in 2011, there has been enormous growth in the development of improved methods in the field of secondary structure prediction (9, 19–24), irregular secondary structure prediction (10, 25–28), protein–ligand interactions (7, 15, 16, 29), DNA/RNA–protein interactions (13, 30, 31), protein crystallization and propensity prediction (32–35), dihedral angle prediction (6, 36–38), surface accessibility prediction (39), Rotamer libraries (8) and others (40–43). These methods have been found to annotate protein structure and function in comparison to earlier methods.…”