Summer Intern, Modeling and Informatics, Merck Co. (06/2018-08/2018)
Project: Developed a genetic-algorithm based molecular evolver as a digital assistant for de novo drug design
- Built a Pipeline Pilot workflow to streamline molecule generator, molecule annotator, and molecule selector for automatic generation of new compounds with desirable properties
- Implemented deep learning models for molecule generation and assessment of synthetic complexity for proposed idea molecules
- Worked closely and effectively with discovery project modelers and applied the molecule evolver to 4 ongoing Merck projects for lead compound discovery & optimization
Summer Intern, Computational Chemistry, Biogen Co. (06/2017-08/2017)
Project: Developed machine learning models to predict P-Glycoprotein substrate for in-house compounds
- Developed 5 single and 3 ensemble classifiers for prediction of P-gp substrates for 3000+ compounds. The best model achieves more than 85% accuracy.
- Developed a workflow to automate feature selection, hyper-parameter tuning, model validation, and model evolution for each learning models.
Pharmacometrician Intern, Center for Clinical Pharmaceutical Sciences, University of Pittsburgh (05/2014-08/2014)
Project: Developed ordinary-differential equation (ODE) mathematical models to study the pathophysiological mechanism of Gestational diabetes mellitus.
Application Scientist Intern, CloudScientific Technology Co. Ltd, Shanghai, China (05/2012-07/2012)
Project: Application scientist of computational chemistry scientific software (MOE/LeadIT/Derek) for drug design, lead compound optimization, toxicity prediction.