Kumud Pant

Decision Tree Classifier for Classification of Plant and Animal Micro RNA’s

Gene expression is regulated by miRNAs or micro RNAs which can be 21-23 nucleotide in length. They are non coding RNAs which control gene expression either by translation repression or mRNA degradation. Plants and animals both contain miRNAs which have been classified by wet lab techniques. These techniques are highly expensive, labour intensive and time consuming. Hence faster and economical computational approaches are needed. In view of above a machine learning model has been developed for classification of plant and animal miRNAs using decision tree classifier. The model has been tested on available data and it gives results with 91% accuracy.

Association Rule Mining to Deduce the Most Frequently Occurring Amino Acid Patterns in HIV

HIV is one of the most dreaded diseases of the century. Throughout the world efforts are underway to develop new vaccines and design new drugs so as to combat this viral menace. In an effort to probe deeper into the functioning of these viruses we present association based rules formulation so as to decipher the most frequently occurring amino acids in these viruses. This is a novel attempt of its kind since we are attempting to find put the most informative association rules using Apriori algorithm implemented through WEKA. The information generated can be of great use to molecular biologists and drug designers since the associated amino acids can be a very good drug targets. Our findings suggest that L-Selenocysteine and L-Pyrrolysine are most frequently associated amino acids in the 4 classes of virulent proteins analyzed for association rules and Cyteine and Arginine show the strongest association in one of the class analyzed i.e. Gp41. Hence these can be potential drug candidates. General Terms Association Rule Mining, HIV, Apriori Algorithm.

Multi Class Classification Approach for Classification of ADAMs, MMPs and Their Subclasses

The MMPs and ADAMs are cell surface proteases which belong to metalloprotease family. They play an important role in skin aging, skin disorders, anticancer therapy and other physiological disorders. Thus there arises the need to understand the relationships among various parameters of these proteins for prediction of their classes, structures and functionality. The computational approaches for prediction of their classes are fast and economical therefore can be used to complement the existing wet lab techniques. Realizing their importance, in this paper an attempt has been made to correlate them with their amino acid composition and predict them with fair accuracy. This is a novel method where ADAMs and MMPs have been classified on the basis of amino acid composition using Support Vector Machine. The SVM has been implemented using Lib SVM package. The method discriminates MMP subfamily from ADAM proteases with Matthews correlation coefficient of 0.98 using amino acid composition. The method is further able to predict three major subclasses or subfamilies of MMPs with an overall Matthews correlation coefficient (MCC) and accuracy of 0.782 and 89.01% respectively using amino acid composition. The performance of the method was evaluated using 5-fold cross-validation where accuracy of 98% was obtained.

Naïve Bayes Classifier for Classification of Plant and Animal miRNA

MicroRNAs (miRNA) are single-stranded RNA molecules of about 21-23 nucleotides in length. MicroRNAs (miRNAs) constitute a large family of non coding RNAs that function to regulate gene expression. Till today wet lab experiments have been used to classify the miRNA of plants and animals. The wet lab techniques are highly expensive, labour intensive and time consuming. Thus there arises a need for computational approach for classification of plants and animal miRNA. These computational approaches are fast and economical as compared to wet lab techniques. In view of above a machine learning models has been developed for classification of plant and animal miRNA using Naive Bayes classifier. The model has been tested on available data and it gives results with 85.71% accuracy.

ProCoS: Protein composition server.

ProCoS is a free online tool for computing different combinations of peptide compositions. It is developed as an applet and a server with a capability to handle multiple FASTA sequences. The generalized algorithm for computing poly-amino acid composition forms the core of ProCoS. It produces output in different formats for easy visualization of results. It also allows composition analysis of sequences in full or in specific parts. Thus, ProCoS is user-friendly, flexible and unique.

A Model for Detection, Classification and Identification of Spam Mails Using Decision Tree Algorithm

Spam mails are unsolicited bulk mails which are meant to fulfill some malicious purpose of the sender. They may cause economical, emotional and time losses to the recipients. Hence there is a need to understand their characteristics and distinguish them from normal in box mails. Decision tree classifier has been trained with the major characteristics of spam mails and results obtained with more then 86.7437% accuracy. This classifier can be a valuable strategy for software developers who are trying to combat this ever growing problem.

SVM Model for Amino Acid Composition Based Prediction of MMPs and ADAMs

The MMPs and ADAMs are cell surface proteases which belong to metalloprotease family. They play an important role in skin aging, skin disorders, anticancer therapy and other physiological disorders. Thus there arises the need to understand the relationships among various parameters of these proteins for prediction of their classes, structures and functionality. The computational approaches for prediction of their classes are fast and economical therefore can be used to complement the existing wet lab techniques. Realizing their importance, in this paper an attempt has been made to correlate them with their amino acid composition and predict them with fair accuracy. This is a novel method where ADAMs and MMPs have been classified on the basis of amino acid composition using Support Vector Machine. The SVM has been implemented using Lib SVM package. The method discriminates MMP subfamily from ADAM proteases with Matthews correlation coefficient of 0.98 using amino acid composition. The performance of the method was evaluated using 5-fold cross-validation where accuracy of 98% was obtained.

Issues on machine learning for prediction of classes among molecular sequences of plants and animals

Nowadays major laboratories of the world are turning towards in-silico experimentation due to their ease, reproducibility and accuracy. The ethical issues concerning wet lab experimentations are also minimal in in-silico experimentations. But before we turn fully towards dry lab simulations it is necessary to understand the discrepancies and bottle necks involved with dry lab experimentations. It is necessary before reporting any result using dry lab simulations to perform in-depth statistical analysis of the data. Keeping same in mind here we are presenting a collaborative effort to correlate findings and results of various machine learning algorithms and checking underlying regressions and mutual dependencies so as to develop an optimal classifier and predictors.

Mining Genomic Patterns in Mycobacterium tuberculosis H37Rv Using a Web Server Tuber-Gene

Mycobacterium tuberculosis (MTB), causative agent of tuberculosis, is one of the most dreaded diseases of the century. It has long been studied by researchers throughout the world using various wet-lab and dry-lab techniques. In this study, we focus on mining useful patterns at genomic level that can be applied for in silico functional characterization of genes from the MTB complex. The model developed on the basis of the patterns found in this study can correctly identify 99.77% of the input genes from the genome of MTB strain H37Rv. The model was tested against four other MTB strains and the homologue M. bovis to further evaluate its generalization capability. The mean prediction accuracy was 85.76%. It was also observed that the GC content remained fairly constant throughout the genome, implicating the absence of any pathogenicity island transferred from other organisms. This study reveals that dinucleotide composition is an efficient functional class discriminator for MTB complex. To facilitate the application of this model, a web server Tuber-Gene has been developed, which can be freely accessed at http://www.bifmanit.org/tb2/.

Dipeptide based SVM model for prediction of CDKs and cyclins

Various combination of both cyclin dependent kinases (CDKs) and cyclin proteins are responsible for progression of cell cycle through various phases like G1, S, G2 and M. CDKs are enzymes with possible role to play in anti cancer therapy. Realizing the importance of both these proteins in various aspects of life a new efficient computational model has been developed using parameters like dipeptide composition for prediction of these proteins. The support vector machine (SVM) package used has been implemented using freely downloadable software LIBsvm. With five fold cross validation accuracy of 99.9644% has been achieved in predicting the two classes using dipeptide composition (DPC). Further the accuracy of test module came out to be 95.6989%.