Phase separation of proteins play key roles in cellular physiology including bacterial division, tumorigenesis etc. Consequently, understanding the molecular forces that drive phase separation has gained considerable attention and several factors including hydrophobicity, protein dynamics, etc., have been implicated in phase separation. Data-driven identification of new phase separating proteins can enable in-depth understanding of cellular physiology and may pave way towards developing novel methods of tackling disease progression. In this work, we exploit the existing wealth of data on phase separating proteins to develop sequence-based machine learning method for prediction of phase separating proteins. We use reduced alphabet schemes based on hydrophobicity and conformational similarity along with distributed representation of protein sequences and biochemical properties as input features to Support Vector Machine (SVM) and Random Forest (RF) machine learning algorithms. We used both curated and balanced dataset for building the models. RF trained on balanced dataset with hydropathy, conformational similarity embeddings and biochemical properties achieved accuracy of 97%. Our work highlights the use of conformational similarity, a feature that reflects amino acid flexibility, and hydrophobicity for predicting phase separating proteins. Use of such "interpretable" features obtained from the ever-growing knowledgebase of phase separation is likely to improve prediction performances further.

Human Serum Albumin (HSA) has been suggested to be an alternate biomarker to the existing Hemoglobin-A1c (HbA1c) marker for glycemic monitoring. Development and usage of HSA as an alternate biomarker requires the identification of glycation sites, or equivalently, glucose-binding pockets. In this work, we combine molecular dynamics simulations of HSA and the state-of-art machine learning method Support Vector Machine (SVM) to predict glucose-binding pockets in HSA. SVM uses the three dimensional arrangement of atoms and their chemical properties to predict glucose-binding ability of a pocket. Feature selection reveals that the arrangement of atoms and their chemical properties within the first 4Å from the centroid of the pocket play an important role in the binding of glucose. With a 10-fold cross validation accuracy of 84 percent, our SVM model reveals seven new potential glucose-binding sites in HSA of which two are exposed only during the dynamics of HSA. The predictions are further corroborated using docking studies. These findings can complement studies directed towards the development of HSA as an alternate biomarker for glycemic monitoring.

Protein function prediction is an important problem in functional genomics. Typically, protein sequences are represented by feature vectors. A major problem of protein datasets that increase the complexity of classification models is their large number of features. The process of drug discovery often involves the use of quantitative structure-activity relationship (QSAR) models to identify chemical structures that could have good inhibitory effects on specific targets and have low toxicity (non-specific activity). QSAR models are regression or classification models used in the chemical and biological sciences. Because of high dimensionality problems, a feature selection problem is imminent. In this study, we thus employ a hybrid Estimation of Distribution Algorithm (EDA) based filter-wrapper methodology to simultaneously extract informative feature subsets and build robust QSAR models. The performance of the algorithm was tested on the benchmark classification challenge datasets obtained from the CoePRa competition platform, developed in 2006. Our results clearly demonstrate the efficacy of a hybrid EDA filter-wrapper algorithm in comparison to the results reported earlier.

Microarray cancer gene expression data comprise of very high dimensions. Reducing the dimensions helps in improving the overall analysis and classification performance. We propose two hybrid techniques, Biogeography - based Optimization - Random Forests (BBO - RF) and BBO - SVM (Support Vector Machines) with gene ranking as a heuristic, for microarray gene expression analysis. This heuristic is obtained from information gain filter ranking procedure. The BBO algorithm generates a population of candidate subset of genes, as part of an ecosystem of habitats, and employs the migration and mutation processes across multiple generations of the population to improve the classification accuracy. The fitness of each gene subset is assessed by the classifiers - SVM and Random Forests. The performances of these hybrid techniques are evaluated on three cancer gene expression datasets retrieved from the Kent Ridge Biomedical datasets collection and the libSVM data repository. Our results demonstrate that genes selected by the proposed techniques yield classification accuracies comparable to previously reported algorithms.

Antimicrobial peptides (AMPs) are gaining popularity as anti-infective agents. Information on sequence features that contribute to target specificity of AMPs will aid in accelerating drug discovery programs involving them. In this study, an algorithm called ClassAMP using Random Forests (RFs) and Support Vector Machines (SVMs) has been developed to predict the propensity of a protein sequence to have antibacterial, antifungal, or antiviral activity. ClassAMP is available at http://www.bicnirrh.res.in/classamp/.

Automatically classifying ECG recordings for malignant ventricular arrhythmia is fraught with several difficulties. Even normal ECG signals exhibit only quasi-periodic nature, and contain various irregularities. The key to more accurate detection is the use of position, and amount of local singularities in the signals.In this paper, we propose a Holder-SVM detection algorithm using a novel hybrid arrangement of binary and multiclass SVMs designed to take care of class imbalance rampant in biomedical signals. As a result, we significantly reduce the number of false negatives - patients falsely classified as normal. We used the MIT-BIH Arrhythmia database for even different arrhythmias. We compare our hybrid SVM with a suitable conventional SVM, and show better results.We also use the new arrangement for features proposed earlier, and demonstrate the gain in accuracy. Our concept of hybrid SVM is applicable to a wide variety of multiclass classification problems.

Extensive research has been done to show that heartbeats are composed of the interaction of many physiological components operating on different time scales, with nonlinear and self-regulating nature. The more direct, and easily accessible manifestation of the heartbeat is the pulse; however, it has not been studied anywhere near as extensively. In this paper, we establish the relevance of the multi-fractal formalism for the arterial pulse, which has long been used as a fundamental tool for diagnosis in the Traditional Indian Medicine, (Ayurveda). The finding of power-law correlations through detrended fluctuation analysis indicates presence of scale-invariant, fractal structures in the pulse. These fractal structures are then further established by self-affine cascades of beat-to-beat fluctuations revealed by wavelet decomposition at different time scales. Finally, we investigate how these pulse dynamics change with age, and disorder. The analytic tools we discuss may be used on a wide range of physiological signals.

Heart rate variability (HRV) provides an estimate of sympathetic and parasympathetic influences on the heart rate. Although HRV has been extensively studied, sustained clinical use is still outstanding. The noninvasive, convenient, and inexpensive arterial pulse originate from heartbeats, but has not been studied in a systematic fashion except in rudimentary ways. In this paper, we present pulse rate variability (PRV) as an alternative to HRV. We give evidence for the detection of disorders in patients using PRV, paving the way for future clinical use.

MOTIVATION: Inclusion body formation has been a major deterrent for overexpression studies since a large number of proteins form insoluble inclusion bodies when overexpressed in Escherichia coli. The formation of inclusion bodies is known to be an outcome of improper protein folding; thus the composition and arrangement of amino acids in the proteins would be a major influencing factor in deciding its aggregation propensity. There is a significant need for a prediction algorithm that would enable the rational identification of both mutants and also the ideal protein candidates for mutations that would confer higher solubility-on-overexpression instead of the presently used trial-and-error procedures. RESULTS: Six physicochemical properties together with residue and dipeptide-compositions have been used to develop a support vector machine-based classifier to predict the overexpression status in E.coli. The prediction accuracy is approximately 72% suggesting that it performs reasonably well in predicting the propensity of a protein to be soluble or to form inclusion bodies. The algorithm could also correctly predict the change in solubility for most of the point mutations reported in literature. This algorithm can be a useful tool in screening protein libraries to identify soluble variants of proteins.

MOTIVATION: Accurate prediction of genes in genomes has always been a challenging task for bioinformaticians and computational biologists. The discovery of existence of distinct scaling relations in coding and non-coding sequences has led to new perspectives in the understanding of the DNA sequences. This has motivated us to exploit the differences in the local singularity distributions for characterization and classification of coding and non-coding sequences. RESULTS: The local singularity density distribution in the coding and non-coding sequences of four genomes was first estimated using the wavelet transform modulus maxima methodology. Support vector machines classifier was then trained with the extracted features. The trained classifier is able to provide an average test accuracy of 97.7%. The local singularity features in a DNA sequence can be exploited for successful identification of coding and non-coding sequences. CONTACT: Available on request from [email protected].