Submit Your Manuscript Online

Prediction of Drug - Target Interaction Using Bipartite Local Models and K-Nearest Neighbor

	International Journal of Scientific Research and Engineering Development (IJSRED)	Full Text:
	Published Issue : Volume-3 Issue-5
	Year of Publication : 2020
	Unique Identification Number : IJSRED-V3I5P47
	Authors : R. Kowsalya, Dr.E.Siva Sankari
	: Click Here

Abstract :

Computational prediction of drug-target interaction is an essential task with various applications in the pharmaceutical industry. The drug-target interaction is being evaluated using biochemical validation of hypothesized drug interaction which is considered time-consuming, laborious and expensive. To overcome the restrictions of the traditional approach there are machine learning based drug-target interactions prediction methods. In this work, the Bipartite Local Model (BLM), one of the most prominent machine learning technique is used in interaction prediction. In particular, BLM with a hubness-aware regression technique and k-nearest neighbor method is used. Similarity between drugs and target is identified using Tanimoto coefficient and Smith Waterman score method respectively. The results are then processed using the BLM to predict the interaction between the drug and the target. In this method, the interaction with multiple set of features like Chemical, genomic and interaction features are predicted. Then, hubness-aware regression is used to rectify the error that may occur while applying BLM. KNN is then applied on to the results obtained through BLM so as to identify the value and understand the nearest of the interaction predicted using BLM. Finally the accuracy of the drug-target interaction is evaluated and also the drug-target pair is identified.