Development of physics-based and data-based molecular simulation methods

Abstract

Recently, molecular simulations have benefited greatly from the development and subsequent application of deep-learning methods. In this talk, we will discuss how machine-learning methods can be combined with enhanced sampling techniques to speed up molecular dynamic simulations. We will then discuss about our recent effort on learning from Alphafold2 to reproduce and improve protein structure prediction AI models. All these methods are implemented in our home-made molecular simulation package, SPONGE, an MD software rewritten using MindSpore and highly compatible with the deep-learning platform. Through these efforts, we try to generate a multi-functional package for structure prediction, molecule and sequence generation, structure evaluation, and dynamics simulation. We will also discuss the possible applications of these methods in physical, chemical and biological problems.

Speaker

本科毕业于四川大学化学系，1996 年在中科院化学所获得硕士学位，2001 年获得加州理工学院博士学位。2001 年－ 2004 年在加州理工学院和哈佛大学做博士后研究。2004 年 -2010 年在美国德克萨斯农工大学（Texas A&M University）化学系任助理教授；2010 年起任北京大学化学与分子工程学院教授，2013 年起同时担任北京大学生物医学前沿创新中心研究员。主要从事生物物理化学/ 理论化学方面的基础研究。现任北京大学理学部副主任，JCTC杂志副主编 。