Our group’s research focuses on combining experimental and computational approaches to explain and predict the behavior of biologically interesting macromolecules. The current research program is aimed at the understanding of the molecular mechanisms and bioinformatic studies of disease-causing genetic variations, protein folding, stability, and binding of ligands, peptides, and DNA. The long-term goal of the proposed research is to elucidate the relations between the sequence, structure, and function of proteins and to uncover the molecular mechanisms underlying different diseases. This will be accomplished by designing simple (yet realistic) models and by developing statistical mechanics theories and bioinformatic tools. Two selected projects are described below.
Experimental and computational Protein/Peptide drug and design
Biological drugs such as proteins, nucleic acids, and sugars are medicinal products extracted from living systems. Unlike small molecular drugs, they are found to be more specific to the intended target, and more effective in interacting with multiple receptors. More importantly, developing biological drugs enjoys much higher success rate than small molecular drugs in every stage of drug development from preclinical to registration. It was estimated that 50% of top 100 selling drugs are biological drugs this year. This percentage is expected to increase further as the gap widens between the number of patents filed for biologics and the number filed for small molecules. However, unlike small molecules, biologic drugs are expensive to manufacture due to low yields and short in shelf-life due to their low stability. The objectives of our research are to improve drug production by codon optimisation, to increase drug stability by protein engineering, and to design peptides or proteins with specific therapeutic effects (inhibitors and vaccines).
Our small-molecule design technique differs from classical docking methods by building on the knowledge of existing ligand-protein interactions. This approach takes advantage of known structural and interaction information of both ligands and protein receptors and decreases significantly the false positive rate. Dr. Zhou’s group is using this approach with Prof Mark von Itstein’s group in developing novel inhibitors for sialidase, the target for antiviral therapy.
Target/off-target Discovery and Drug Repurposing
The targets for many drugs are unknown while drugs with known targets may interact with unintended targets (off-targets) that could lead to side effects or beneficial use for other diseases (drug repurposing). We are developing an approach that systematically scans over whole proteomes of human and pathogenetic species for possible targets by using largest ligand-target datasets. We have been working with BioDiem Inc. to identify the targets of an anti-bacteria agent with success.
Novel bioinformatics methods for prioritizing disease-causing genetic variations
One key element of personalized medicine is having accurate genetic tests that separate diseasing-causing/susceptible human genetic variants from neutral ones. Based on the Human Genome Mutation Database (HGMD), single nucleotide variations (SNVs), the largest class of genetic variations, are responsible for over 50% of known Mendelian diseases. This is followed by micro-insertions and deletions (INDELs, less than 20 nt), responsible for 24% known Mendelian diseases. So far, million of SNVs and micro-INDELs have been observed in humans. Obviously, it is not practical to examine the biological functions of each variant. Thus, there is a critical need for effective tools for distinguishing the likely disease-causing variants from the ones that are functionally neutral. Our approach is to integrate sequence, structure and functional information of genes and its regulatory elements for more accurate discrimination of gene variations. Our long-term goal is to build an integrated tool for predicting susceptibility for specific disease/phenotype from all types of genetic variations.