Bharath received a BA and BS from UC Berkeley in EECS and Mathematics and was valedictorian of his graduating class in mathematics. He did his PhD in computer science at Stanford University where he studied the application of deep-learning to problems in drug-discovery. At Stanford, Bharath created the deepchem.io open-source project to grow the deep drug discovery open source community, co-created the moleculenet.ai benchmark suite to facilitate development of molecular algorithms, and more. Bharath’s graduate education was supported by a Hertz Fellowship, the most selective graduate fellowship in the sciences. After his PhD, Bharath co-founded Computable a startup that built better tools for collaborative dataset management. Bharath is currently working actively on growing the DeepChem community and on exploring a few early projects still in stealth.
Bharath is the lead author of “TensorFlow for Deep Learning: From Linear Regression to Reinforcement Learning”, a developer’s introduction to modern machine learning, with O’Reilly Media, and the lead author of “Deep Learning for the Life Sciences”
Books
- “TensorFlow for Deep Learning: From Linear Regression to Reinforcement Learning”
- “Deep learning for the life sciences”
Essays
- A PhD in Snapshots
- The Advent of Huang’s Law
- Why Blockchain Could (One Day) Topple Google
- Learning to Learn
- What is Ethereum?
- Liquidity in Drug Discovery
- The Innovator’s Open Source Dilemma
- Building The Open Source Drug Discovery Ecosystem
- What Can’t Deep Learning Do?
- Machine Learning With Small Data
- Software Patents Are Obsolete in the Age of AI
- தமிழில் விஞ்ஞானத்தை வளர்ப்போம்! (Science for all languages!)
- A Short Overview of Drug Discovery
- Why Should Drug Discovery Be Open Source?
- Software Is Eating Science
- Can AI Believe in God? A Parable about Diversity
- Learning Models of Disease
- The Ferocious Complexity of the Cell
- Why Antibiotics Are Hard
- Machine Learning For Scientific Datasets
- Can Drugs Be Developed Like Open Source Projects
Papers
- Massively multitask networks for drug discovery
- Low data drug discovery with one-shot learning
- NVMKV: A Scalable and Lightweight Flash Aware Key-Value Store.
- MoleculeNet: a benchmark for molecular machine learning
- Atomic convolutional networks for predicting protein-ligand binding affinity
- Computational modeling of β-secretase 1 (BACE-1) inhibitors using ligand based approaches
- Understanding protein dynamics with L1-regularized reversible hidden Markov models
- Is multitask deep learning practical for pharma?
- Retrosynthetic reaction prediction using neural sequence-to-sequence models
- The extended parameter filter
- Dynamic scaled sampling for deterministic constraints
- Spatial Graph Convolutions for Drug Discovery
- Learning Protein Dynamics with Metastable Switching Systems
- A guide to deep learning in healthcare
- Solving the RNA design problem with reinforcement learning
- PotentialNet for molecular property prediction
- Secure Computation in Decentralized Data Markets
- AMPL: A Data-Driven Modeling Pipeline for Drug Discovery
- ChemBERTa: Large-Scale Self-Supervised Pretraining for Molecular Property Prediction
Patents
- Non-volatile key-value store
- Conditional iteration for a non-volatile device
- Interpretability-based machine learning adjustment during production
- Determining validity of machine learning algorithms for datasets
- Detecting suitability of machine learning models for datasets
- Systems and Methods for Spatial Graph Convolutions with Applications to Drug Discovery and Molecular Simulation
- Machine learning abstraction
Technical Blog Posts
- Mixture Descriptors toward the Development of Quantitative Structure–Property Relationship Models for the Flash Points of Organic Mixtures
- Lattice convolutional neural network modeling of adsorbate coverage effects
- Ultra-large library docking for discovering new chemotypes
- Deciphering interaction fingerprints from protein molecular surfaces
- Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
- Unsupervised rational protein engineering with sequence only deep representation learning
- Induced Pluripotency as a Benchmark for Differentiable Cell Simulators
- Making Deep Learning Useful for Small Life Science Datasets
- Rethinking the Use of NLP Methods in the Life Sciences
Contact
X.Y@gmail.com, sub X=bharath, Y=ramsundar