Deeptak Verma, PhD

Research

Directed evolution

Directed evolution is a powerful cyclic technique that mimics natural selection of proteins towards a property of interest. Some of the applications include improving protein stability, enhancing enzyme activity, improving antibody affinity, etc. My research on directed evolution focuses on employing computational techniques to design smart protein libraries that can accelerate the protein design efforts, specifically enzymes. Click here for related publication.

Machine learning

There has been an increased interest in using machine learning to assist protein redesign, since prediction models can be used to virtually screen large number of novel sequences. A good prediction model is dependent on many factors, such as, data type, data filtering, feature extraction and the machine learning algorithm. My research explores different methods and strategies to help build better machine learning models in the area of protein reengineering. Click here for related publication.

Engineering therapeutic proteins

Computational protein engineering focuses on designing beneficial proteins by modeling sequence-structure-function relationships and optimizing them with desired properties. My research focuses on designing safe therapeutic proteins for treatment of infectious diseases (eg: MRSA), and also designing focused (HIV, influenza and dengue) immunogenic proteins to elicit native-like immune responses. Click here for related publication.

Sequence-structure-function relationships

Proteins evolve and adapt to newer functions by introducing mutations, while maintaining their structural properties. Assessment of physiochemical and biophysical properties both at the sequence and structure level can provide insights on evolution of such functions. My research on beta-lactamase proteins (that are responsible for antibiotic resistance) provides insights on the relationships between sequence evolution, structure properties and function conservation. Click here for related publication.

Computer-aided drug design

Rational drug design aims at developing small molecule drugs that can modulate the function of a given protein target. Many different analyses are performed to identify putative successful drug molecules, such as, 3D-QSAR studies, binding free energy assessments, molecular docking, ADME properties, etc. My research in this domain has focused on developing drugs against HIV-1 reverse transcriptase at the p66 domain and SNAI1 proteins for the treatment of cancer. Click here for related publication.

Protein flexibility and stability

Proteins are not static molecules! Computational characterization of the changes in stability and flexibility of proteins due to mutations can provide biophysical insights. Computational models can be developed for reproducing experimental stability trends, whereas protein dynamics analysis can identify regions with high flexibility/rigidity. My assessment of dynamical properties on lysozyme proteins suggests that the changes due to mutations can be frequent, large and long-ranged. Click here for related publication.

Big data analysis & bio-visualization

A picture is worth a thousand words! Presentation, analysis and interpretation of large amounts of biological data becomes easier when presented in a simple and clear way. Combined with clustering and classification tools, bio-visualization can help with feature identification and comparative studies, thus streamlining pharmaceutical research. In this collaborative research work, I have helped in the development of various bio-visualization tools. Click here for related publication.