Paper Image

Autoregressive diffusion modeling for structure-based drug design

Published on:

2 April 2024

Primary Category:

Machine Learning

Paper Authors:

Xinze Li,

Penglei Wang,

Tianfan Fu,

Wenhao Gao,

Chengtao Li,

Leilei Shi,

Junhong Liu

Bullets

Key Details

Proposes AUTODIFF, a diffusion-based fragment-wise autoregressive model for structure-based drug design

Introduces conformal motif design strategy to preserve local structure conformation

Generates molecules motif-by-motif using diffusion modeling

Evaluated on CrossDocked2020 dataset

Shows improved performance in generating realistic molecules with valid structure/conformation and high binding affinity

AI generated summary

Autoregressive diffusion modeling for structure-based drug design

This paper proposes AUTODIFF, a novel diffusion-based fragment-wise autoregressive generation model for structure-based drug design. It introduces a conformal motif design strategy that preserves local structure conformation, and uses diffusion modeling to generate molecules motif-by-motif. The model is evaluated on CrossDocked2020 and shows improved performance in generating realistic molecules with valid structures/conformations and high binding affinity.

Answers from this paper

Comments

No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up