1. Structure based pharmacology for cancer and immunology
Using atomic data on drug targets to develop new cancer and immunology drugs. We are collaborating with Eureka Therapeutics and Dr. David Scheinberg at Memorial Sloan-Kettering Cancer Center to structurally characterize antibody therapeutics to better understand their mechanism of action and help engineer superior next generation antibodies against such targets as the ROR2 tyrosine kinase receptor and HLA-Wilms Tumor 1 peptide complex, which has recently demonstrated dramatic efficacy in leukemia in mice. A second focus is anti-estrogen drugs, which are important for treating breast cancer. We are investigating new mechanisms of targeting aromatase, the enzyme responsible for estrogen biosynthesis, for breast cancer. We are developing immunomodulatory drugs that inhibitor or activate the key immune switch, RORgamma, for the treatment of autoimmune disorders and cancer. We have also recently determined the crystal structure of a new botulinum toxin antidote for biodefense applications in collaboration with Hawaii Biotech. Our X-ray experiments are performed at state of the art, particle accelerator facilities such as the Stanford Synchrotron Radiation Lightsource and the Advanced Light Source.
2. Hydrogen atoms in protein structure and function
Detecting invisible hydrogen atoms in proteins. We are developing a new computational method, HyPO (Hydrogen atom Prediction and Observation) for analyzing protein X-ray crystallography maps to detect hydrogen atoms. Hydrogen atoms, having only one electron, scatter X-rays very weakly and are often invisible in X-ray maps. HyPO locates hydrogen atoms, which play critical roles in protein function such as enzyme mechanism and ligand binding. We are developing HyPO to work with x-ray crystallography maps of modest resolution and weak neutron crystallography maps. HyPO predictions will be tested experimentally by x-ray and neutron crystallography. Neutron crystallography is performed in collaboration with colleagues at Oak Ridge National Laboratory.
3. Applications of machine learning/artificial intelligence to chemistry and drug design
AI-powered drug design. We are developing machine learning methods to generate new molecules for drug design. Machine learning provides new strategies for sampling the massive diversity of chemical space to find molecules with drug-like properties. We are currently focusing on identifying new inhibitors for K-Ras, a famously "undruggable" target commonly mutated in pancreatic and lung cancer, as well as estrogen receptor, a primary drug target for breast cancer. This is a collaboration with Prof. Alan Aspuru-Guzik's lab at the University of Toronto. We are also interested in developing AI methods from computer vision/image processing for analyzing molecular images from crystallography, electron microscopy, and molecular dynamics simulations.
4. Mechanisms of photoactive proteins
Shining light on photoactive proteins. Photoactive proteins play central roles in state-of-the-art microscopy. We seek to understand physical mechanisms of fluorescent proteins to facilitate their applications to microscopy and as biosensors. This work is done in collaboration with Prof. Michael Lin of Stanford, Prof. Hui-wang Ai of the University of Virginia, and Prof. Jun Chu of the Shenzhen Institute of Advanced Technologies. we are helping to develop new fluorescent proteins with improved spectral properties, that can act as biosensors and optical switches. Through crystallographic and computational methods, we are developing theories of chemical mechanisms that can guide the development of fluorescent proteins with improved brightness and spectral properties.
5. Open Source Malaria
Structure-based drug discovery for malaria. Malaria continues to be among the deadliest infectious diseases in the world. Drug resistance continues to grow with few new drug candidates on the horizon. Market failures in the industrial sector, combined with lack of funding and cultural challenges in the academic sector has frustrated drug discovery efforts. We are part of a new experiment in how research is structured and conducted. As part of the Open Source Malaria consortium, we are inspired by the success of the open source software model. We subscribe to a completely transparent model, where all data and methods are shared, and participation is open to all. Our lab contributes to computational chemistry, structure-based drug design, and identification of drug targets. We are particularly interested in protein kinases as potential drug targets.