Anirudh Tiwari  received his MS-Dual Degree in Computational Natural Sciences (CNS). His  research work was supervised by Dr. Nita Parekh and thesis reviewed by Dr. Deva Priyakumar  and  Dr. Bhaswar Ghosh.

Here’s a summary of  Anirudh’s M.S thesis, Developing an Automated Method to Identify Structural Domains in Proteins as explained by him:  

Proteins are the building blocks of all living organisms. Domains are subunits in  a protein that are critical in determining the structure and function. The focus of our study was to develop an automated method to identify the structural domains in a protein. We modelled protein as a network and extracted properties that were used as input features to a supervised machine learning algorithm (SVM) to identify the number of domains. Further, we used an unsupervised machine learning technique (k-means) to identify the domain boundaries. We benchmarked our algorithm against 4 state-of-the-art domain identification methods and observed that our algorithm performed at par in identifying domains.

The two images depict the algorithm workflow and the result of our algorithm depicting identifying the number of domains as 2 (shown in red and green).