Predicting Protein Structural Features

Proteins play a vital role in the biological activities of all living species. The current experimental methods of protein structure determination are complicated, time-consuming, and expensive. On the other hand, the sequencing of proteins is fast, simple, and relatively less expensive. Thus, the gap between the number of known sequences and the determined structures is growing, and is expected to keep expanding. In contrast, computational approaches that can generate three-dimensional protein models with high resolution are attractive, due to their broad economic and scientific impacts. Accurately predicting protein structural features, such as secondary structures, disulfide bonds, solvent accessibility, flexibility, disorder, residue contact, and torsion angles is a critical intermediate step stone to obtain correct three-dimensional models ultimately.

In this project, a set of machine learning approaches are utilized for improving the accuracy of predicting protein structural features. 

I have focused on improving the predictions of proteins’ structural features, such as predicting the secondary structures, disulfide bonds, solvent accessibility, and residue contacts. Those structural features represent the intermediate steps along the way towards protein tertiary structure prediction. I have also focused on proposing efficient computational approaches, in protein modeling, suitable for multi-core and many-core architectures.

The goal is to develop new systems for protein modeling that use machine learning methods and current high performance computing resources to efficiently predict proteins’ structural features. The outcomes of such systems will increase our understanding of the biological processes leading into better knowledge of protein structures and functions.

Efficient protein modeling tools can eventually lead to the manufacturing of additional drugs to fight against conditions like Alzheimer and Mad-Cow diseases. However, many details remain unknown when considering the nanoscale interactions among the atoms that constitute protein molecules. That makes protein modeling a challenging field and a major research effort in computational biology