University Oral Exam: Molecular Machine Learning with DeepChem, Bharath Ramsundar, Advised by Vijay Pande

University Oral Exam<br><br>Title:&nbsp;Molecular Machine Learning with DeepChem&nbsp;&nbsp;<br>Speaker:&nbsp;Bharath Ramsundar,&nbsp;&nbsp;Advised by Vijay Pande&nbsp;&nbsp;<br>Date: December 12, 2017<br>Time: 3:15pm<br>Location: Clark, Room S360<br><br>Abstract:&nbsp;<br>Machine learning has widely been applied to image, video, and<br>speech datasets, but has not yet achieved broad penetration into<br>chemistry, materials science, or other molecular design applications.<br>However, over the last few years, machine learning and deep learning have<br>achieved notable successes in predicting properties of molecular systems.<br>In this thesis, I present a series of deep learning algorithms that<br>demonstrate strong predictive improvements across a wide range of<br>biochemical tasks such as assay activity modeling, toxicity prediction,<br>protein-ligand binding affinity calculation, and chemical retrosynthesis.<br>In addition to these algorithmic improvements, I introduce the<br>comprehensive benchmark suite MoleculeNet for molecular machine learning<br>algorithms (<a href="https://moleculenet.ai">https://moleculenet.ai</a>) and demonstrate how the technology of<br>one-shot learning can be used for drug discovery applications. The work<br>presented in this thesis culminated in my design and construction of<br>DeepChem (<a href="https://deepchem.io">https://deepchem.io</a>), an open source package for molecular<br>machine learning, which has achieved broad adoption among biotech<br>startups, pharmaceutical companies, and research groups. DeepChem has<br>attracted a thriving community of open source developers and looks to<br>continue growing and expanding as a vibrant research tool.<br>

Date: 
Tuesday, December 12, 2017 - 3:15pm to 4:15pm
location: 
Clark, Room S360