报告摘要

Protein-protein interactions (PPIs) are critical to various biochemical and biological processes. PPI sites and the compounds that target them possess distinct physicochemical properties compared to traditional binding pockets and drugs, making the generation of novel compounds for PPIs or the identification of PPI modulators a challenging task. Additionally, exploring the conformational landscape of protein-protein complexes is essential for understanding their functional dynamics, stability, interactions, and for informing drug design. To address these challenges, we propose four advanced frameworks: iPPIGAN, a deep learning model for generating novel drug-likePPImolecules; GENiPPI, a moleculargeneration framework focused on PPI interfaces; MultiPPIMI, a deep sequence-based learning framework for predicting interactions between specific PPI targets and modulators; and AlphaPPImd, a Transformer-based generative neural network designed to explore the conformational ensembles of protein-protein complex.

报告人简介

王建民，延世大学在读直博生。此前，2014年获得中南大学湘雅药学院学士学位，并就职于多家医药机构六年。目前研究方向主要集中在人工智能用于蛋白-蛋白相互作用调节剂设计和发现以及分子动力学与深度学习的融合。目前在J. Chem. Theory Comput.,J. Chem. Inf. Model.，Brief. Bioinform.，Comput. Struct. Biotechnol. J.等SCI杂志上发表多篇学木论文。

报告时间：2024年9月7日上午8:30-12:00

报告地点：北衡楼1421