Learning objectives

• Therapeutics Commons (TDC-2) presents a collection of datasets, tools, models, and benchmarks integrating cell-type-specific contextual features with machine learning tasks across the range of therapeutics
• Understand the tasks across contextual AI in therapeutics at single-cell resolution in TDC-2
• Understand the advancements and challenges in machine learning and biology that drove the implementation of TDC-2

Speaker biography

Alejandro is a Pre-Doctoral Research Associate at Harvard Medical School’s Department of Biomedical Informatics, advised by Professor Marinka Zitnik. He completed his S.B. at MIT, where he was co-advised by Justin Solomon (MIT CSAIL) and Moon Duchin (Tufts Math) to develop metric geometry models and data science algorithms for assessing fairness in political redistricting as part of an informal collective called the Metric Geometry and Gerrymandering Group, now the MGGG Redistricting Lab (mggg.org). Before starting at HMS, he was a software engineer at Pinterest focused on data engineering and machine learning. Alejandro’s research at HMS lies at the intersection of single-cell biology, foundation models, and DrugAI. He’s led the development of the Therapeutic Commons (tdcommons.ai), focusing on single-cell foundation models and contextual AI. TDC’s usage has reached 30K monthly users since the release of TDC-2 and the team has been invited to spotlight presentations at both MoML2024 at Mila and AIDrugX at NeurIPS2024.