Bhattacharya lab ranks in top 10 of protein structure prediction challenge
Published: Feb 12, 2019 5:57 AM
By Sylvia Masango
A research team led by Debswapna Bhattacharya, assistant professor of computer science and software engineering, placed in the top 10 of an international protein structure prediction.
The Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction, considered the Olympic Games in protein structure prediction, challenges researchers to predict the 3-D structure of protein molecules. Bhattacharya’s team placed No. 9 out of 98 groups from around the world in academia and the industry in the regular category.
“Protein structure prediction is a very important yet highly challenging open problem in structural bioinformatics.” said Bhattacharya, whose team included computer science and software engineering graduate students Md Hossain Shuvo and Rahul Alapati.
“I feel very fortunate to have had the opportunity to establish my research laboratory at Auburn and work closely alongside our students in this challenging topic. We developed several innovative methods by employing powerful artificial intelligence methods, based on deep learning, coupled with stochastic optimization algorithms to tackle this problem. I am pleased to see our group excelled in multiple categories in this highly competitive venue. With this success, Auburn University has gained prominence in the international protein modeling community.”
From May through August, CASP organizers sent protein sequences with unknown structures to research groups around the world. Participating groups were required to submit their computationally predicted structures within a limited timeframe and independent experts evaluated the submissions. The goal was to complete this task as quickly as possible as it would ultimately decrease the time it would take to develop life-saving drugs from years to hours.
CASP is a community-wide experiment to identify and progress the state of the art in modeling protein structure from amino acid sequence. Every two years, participants are invited to submit models for a set of proteins for which the experimental structures are not yet public. Assessments and results are published in a special issue of the journal PROTEINS.
Media Contact: , chris.anthony@auburn.edu, 334.844.3447
The Bhattacharya lab’s prediction for CASP13 protein target T1022s2-D1, evaluated to be the most accurate among all groups.