Paperpile

Referenced Papers (4)

Atomwise

M. Béczi

NeurIPS ML BDD Bioxi. Vol. 10

"This citation is the source describing the AtomNet model and the large-scale dataset used for its training."

Referenced at: 03:05

AI is a viable alternative to high throughput screening: a 318-target study

Izhar Wallach, Denzil Bernard, Kong Nguyen, Gregory Ho, Adrian Morrison, A Stecula

Sci. Rep., 2024

"This paper details the results of a massive, 318-target study validating the AtomNet platform's ability to successfully identify first-in-class hits, even for targets with no prior training data."

Referenced at: 05:37

An efficient graph generative model for navigating ultra-large combinatorial synthesis libraries

A. Peddinti, P. Ginis, C. Chang, B. Anderson, H. van den Bedem

Advances in Neural Information Processing Systems, 35, pp.8731-8743.

"This paper describes the foundational CSLVAE method developed by the speaker's team, which uses a variational autoencoder to efficiently generate synthesizable molecules from massive combinatorial libraries."

Referenced at: 11:50

NGT: Generative AI with synthesizability guarantees discovers MC2R inhibitors from a Tera-scale virtual screen

Saulo H P de Oliveira, Aryan Pedawi, Victor Kenyon, Henry van den Bedem

J. Med. Chem., 2024

"This paper, published the week of the talk, serves as a case study for the successful application of the NGT generative AI method to discover novel, synthesizable inhibitors for the challenging MC2R target."

Referenced at: 14:21