Main responsibility:

Develop innovative models to predict drug bioactivity, ADMET, etc using advanced machine learning or deep learning approaches. Develop predictive models for allosteric modulator screening and optimization. Develop innovative generalizable models to support biologics drug discovery including but not limited to antibody, TCR and mRNA etc. Increase model accuracy, applicability, and generalization by improving or developing novel molecular featurization, model algorithm and architecture, etc. Enhance the interpretation of machine learning predictions. Collaborate with computational chemists, biologists, data scientist, medicinal chemists, and other research scientists to support drug discovery projects, evaluate model performance, and interactively improve existing methods. Develop an automated workflow to assist expert decision making by integrating high equality models or state-of-art methods. Stay informed on the latest novel methodologies, tools, and applications.

Qualifications

PhD in computer science, data science, computational chemistry or biology, mathematics, statistics, physics, engineering, or related disciplines. Experience with biologics drug related methodology development is preferred. Experience with allosteric pocket or modulator related method development is a plus. Experience with AI-assisted protein conformational sampling or analysis is preferred. Expertise in at least one common programming language used in scientific computing and machine learning such as python. Experience with at least one common deep learning framework such as PyTorch, Tensorflow, Keras, Scikit-learn etc. Good understanding of statistics and underling machine/deep learning methodologies. Hands-on experience of methodology or software development with tractable open-source contributions or scientific publications. Strong verbal and written communication skills and ability to work independently and cooperatively. Experience with one of the cheminformatics toolkits (RDKit, Schrodinger suite, etc) is desirable.