[{"data":1,"prerenderedAt":453},["ShallowReactive",2],{"content-query-ltaLLdChXA":3},{"_path":4,"_dir":5,"_draft":6,"_partial":6,"_locale":7,"title":8,"description":9,"date":10,"cover":11,"type":12,"body":13,"_type":447,"_id":448,"_source":449,"_file":450,"_stem":451,"_extension":452},"/news/zh/2854","zh",false,"","昇思MindSpore联合高毅勤团队在JCTC发表的综述文章荣获编辑良择：人工智能增强的分子模拟","近日，北京大学、深圳湾实验室、昌平实验室高毅勤教授团队联合华为昇思 MindSpore 团队在 Journal of Chemical Theory and Computation 上发表的关于人工智能增强分子模拟的综述文章被选为Editor's Choice。","2023-10-26","https://obs-mindspore-file.obs.cn-north-4.myhuaweicloud.com/file/2023/11/10/493b1ee1f6044531a2eade9a5be9c76e.png","news",{"type":14,"children":15,"toc":444},"root",[16,24,48,56,64,76,86,91,99,104,109,114,119,126,135,143,148,155,163,168,176,181,191,201,211,229,237,247,257,270,277,287,298,306,314,324,334,339,348,353,361,371,384,392,400,405,410,420,433],{"type":17,"tag":18,"props":19,"children":21},"element","h1",{"id":20},"昇思mindspore联合高毅勤团队在jctc发表的综述文章荣获编辑良择人工智能增强的分子模拟",[22],{"type":23,"value":8},"text",{"type":17,"tag":25,"props":26,"children":27},"p",{},[28,30,46],{"type":23,"value":29},"近日，",{"type":17,"tag":31,"props":32,"children":33},"strong",{},[34,36,44],{"type":23,"value":35},"北京大学、深圳湾实验室、昌平实验室高毅勤教授团队联合华为昇思 MindSpore 团队在 Journal of Chemical Theory and Computation 上发表的关于人工智能增强分子模拟的综述",{"type":17,"tag":31,"props":37,"children":38},{},[39],{"type":17,"tag":31,"props":40,"children":41},{},[42],{"type":23,"value":43},"文章被选为E",{"type":23,"value":45},"ditor's Choice",{"type":23,"value":47},"。",{"type":17,"tag":25,"props":49,"children":50},{},[51],{"type":17,"tag":31,"props":52,"children":53},{},[54],{"type":23,"value":55},"英文原题：",{"type":17,"tag":25,"props":57,"children":58},{},[59],{"type":17,"tag":31,"props":60,"children":61},{},[62],{"type":23,"value":63},"Artificial Intelligence Enhanced Molecular Simulations",{"type":17,"tag":25,"props":65,"children":66},{},[67],{"type":17,"tag":31,"props":68,"children":69},{},[70],{"type":17,"tag":71,"props":72,"children":75},"img",{"alt":73,"src":74},"image.png","https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110055815.56106750370887070489457317540282:50541109072105:2400:822155CF34B4B0C9D4E285E60DE3DB50235DC17847F0CFAC520C393BA85CC34F.png",[],{"type":17,"tag":25,"props":77,"children":78},{},[79,84],{"type":17,"tag":31,"props":80,"children":81},{},[82],{"type":23,"value":83},"通讯作者:",{"type":23,"value":85}," 高毅勤，北京大学、深圳湾实验室、昌平实验室；杨奕，深圳湾实验室",{"type":17,"tag":25,"props":87,"children":88},{},[89],{"type":23,"value":90},"**作者：**Jun Zhang (张骏), Dechin Chen (陈迪青), Yijie Xia (夏义杰), Yu-Peng Huang (黄渝鹏), Xiaohan Lin (林潇涵), Xu Han (韩旭), Ningxi Ni (倪宁曦), Zidong Wang (王紫东), Fan Yu (于璠), Lijiang Yang (杨立江), Yi Isaac Yang* (杨奕), Yi Qin Gao* (高毅勤)",{"type":17,"tag":25,"props":92,"children":93},{},[94],{"type":17,"tag":31,"props":95,"children":96},{},[97],{"type":23,"value":98},"背景介绍",{"type":17,"tag":25,"props":100,"children":101},{},[102],{"type":23,"value":103},"分子模拟是化学、生物、物理、材料科学等领域的重要研究工具，但当前主流的分子模拟软件往往诞生于几十年前，在很多地方都难以适应当今IT技术的发展。",{"type":17,"tag":25,"props":105,"children":106},{},[107],{"type":23,"value":108},"一方面传统分子模拟软件在支持新设备（如 GPU）或者大规模并行计算时，往往需要对程序进行大幅改动。而另一方面，传统分子模拟采取的暴力计算难以覆盖更大的时间和空间尺度，而基于机器学习的增强采样技术和多尺度建模策略可以突破这一限制。",{"type":17,"tag":25,"props":110,"children":111},{},[112],{"type":23,"value":113},"但目前的 AI 框架多以 Python 语言为前端，而目前大部分的分子模拟软件以 C/C++和 FORTRAN 语言编写的，导致 AI 算法难以集成在分子模拟软件中。因此只有开发基于 AI 框架的分子模拟软件，才能使AI技术充分在分子模拟中发挥能力。",{"type":17,"tag":25,"props":115,"children":116},{},[117],{"type":23,"value":118},"本文将探讨分子模拟与人工智能对接的潜力和可行性，介绍由 AI 增强分子模拟所实现的应用案例，并展望 AI 增强分子模拟未来发展的新趋势。",{"type":17,"tag":25,"props":120,"children":121},{},[122],{"type":17,"tag":71,"props":123,"children":125},{"alt":73,"src":124},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110055901.62705459037268787903939100164532:50541109072105:2400:4F9816CE097D009BD36C4E4AB80EE4F16AA503FAE3B6075D224DD55D32F3AE88.png",[],{"type":17,"tag":25,"props":127,"children":128},{},[129],{"type":17,"tag":130,"props":131,"children":132},"em",{},[133],{"type":23,"value":134},"图1. 人工智能增强的分子模拟",{"type":17,"tag":25,"props":136,"children":137},{},[138],{"type":17,"tag":31,"props":139,"children":140},{},[141],{"type":23,"value":142},"文章亮点",{"type":17,"tag":25,"props":144,"children":145},{},[146],{"type":23,"value":147},"文章提出了一种“类AI”的分子动力学模拟程序架构，可以如执行 AI 训练那样运行分子动力学（molecular dynamics, MD）模拟程序。",{"type":17,"tag":25,"props":149,"children":150},{},[151],{"type":17,"tag":71,"props":152,"children":154},{"alt":73,"src":153},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110055913.32768335248738847526314320036578:50541109072105:2400:4B4727E2EF40F73210B5CADA0AB0FD5FDDD346026C709B1581709D23651D84B3.png",[],{"type":17,"tag":25,"props":156,"children":157},{},[158],{"type":17,"tag":130,"props":159,"children":160},{},[161],{"type":23,"value":162},"图2. 人工智能训练（上）与分子动力学模拟（下）过程的对比",{"type":17,"tag":25,"props":164,"children":165},{},[166],{"type":23,"value":167},"在该架构下，MD 模拟被看作一种特殊的 AI 训练过程，原子坐标相当于神经网络参数，势能函数相当于作为优化目标的损失函数，而 MD 积分器则相当于一种特殊的AI优化器。根据这种相似性逻辑，分子模拟程序可以自然而然地在AI框架中重构，高毅勤教授团队因此基于全场景 AI 融合框架昇思 MindSpore，开发了 AI 原生分子模拟程序 MindSpore SPONGE。该框架可以充分发挥昇思 MindSpore 作为 AI 框架的强大能力，提供了可微编程、高通量计算和自动硬件迁移等非常有吸引力的功能。",{"type":17,"tag":25,"props":169,"children":170},{},[171],{"type":17,"tag":31,"props":172,"children":173},{},[174],{"type":23,"value":175},"总结展望",{"type":17,"tag":25,"props":177,"children":178},{},[179],{"type":23,"value":180},"人工智能增强的分子模拟是一个全新的交叉学科研究领域，它结合了AI数据驱动和分子模拟物理驱动的优势，并催生出一系列有趣的新技术、新问题和新挑战。我们认为，分子模拟通过结合和适配最先进的 AI 技术，可以构建全新的基础设施，并产生更多独特又令人兴奋的分子模拟全新研究范式。为此，则需要完善以下几个方面：",{"type":17,"tag":25,"props":182,"children":183},{},[184,189],{"type":17,"tag":31,"props":185,"children":186},{},[187],{"type":23,"value":188},"1)",{"type":23,"value":190}," 科学计算算子：AI 框架应支持更多的科学计算函数，以完善和提升AI增强分子模拟程序的功能和性能。",{"type":17,"tag":25,"props":192,"children":193},{},[194,199],{"type":17,"tag":31,"props":195,"children":196},{},[197],{"type":23,"value":198},"2)",{"type":23,"value":200}," 可微模拟和元优化：AI 框架需要支持更高效的反向传播方法，以充分发挥可微模拟和元优化在分子模拟中的潜力。",{"type":17,"tag":25,"props":202,"children":203},{},[204,209],{"type":17,"tag":31,"props":205,"children":206},{},[207],{"type":23,"value":208},"3)",{"type":23,"value":210}," 数据驱动与物理驱动相结合：AI 在分子模拟中的应用平衡数据拟合与物理规则，以避免过拟合并提高模型泛化性。我们希望这些问题在未来得到更多的关注，从而推动人工智能增强分子模拟领域的发展。",{"type":17,"tag":25,"props":212,"children":213},{},[214,216,221,223,228],{"type":23,"value":215},"相关论文发表在 ",{"type":17,"tag":130,"props":217,"children":218},{},[219],{"type":23,"value":220},"Journal of Chemical Theory and Computation",{"type":23,"value":222}," 上，",{"type":17,"tag":31,"props":224,"children":225},{},[226],{"type":23,"value":227},"昌平实验室张骏研究员为文章的第一作者， 北京大学、深圳湾实验室、昌平实验室高毅勤教授与深圳湾实验室杨奕副研究员为通讯作者",{"type":23,"value":47},{"type":17,"tag":25,"props":230,"children":231},{},[232],{"type":17,"tag":31,"props":233,"children":234},{},[235],{"type":23,"value":236},"通讯作者信息",{"type":17,"tag":25,"props":238,"children":239},{},[240],{"type":17,"tag":31,"props":241,"children":242},{},[243],{"type":17,"tag":71,"props":244,"children":246},{"alt":73,"src":245},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110055944.81175567311187872369364027609814:50541109072105:2400:111332E5B7DE4F0232566F7C00F137249289C0D8A5588979FFD799898126F393.png",[],{"type":17,"tag":25,"props":248,"children":249},{},[250,255],{"type":17,"tag":31,"props":251,"children":252},{},[253],{"type":23,"value":254},"高毅勤博士",{"type":23,"value":256},"，北京大学教授，深圳湾实验室资深研究员，昌平实验室领衔科学家。1993年毕业于四川大学，1996年于中国科学院化学研究所获得硕士学位，2001年于美国加州理工学院获得博士学位。随后在美国加州理工学院、美国哈佛大学从事博士后研究，2004-2010年在美国德克萨斯 A&M 大学担任助理教授，2010年加入北京大学担任教授。曾获华为昇腾 MVP 称号，2016年亚太理论与计算化学家协会 Pople Medal，2014年 Promising Scientist Prize of CMOA、日本化学会 Distinguished Lectureship Award 等多个奖项。",{"type":17,"tag":25,"props":258,"children":259},{},[260,262],{"type":23,"value":261},"课题组主页：",{"type":17,"tag":263,"props":264,"children":268},"a",{"href":265,"rel":266},"https://www.chem.pku.edu.cn/gaoyq/",[267],"nofollow",[269],{"type":23,"value":265},{"type":17,"tag":25,"props":271,"children":272},{},[273],{"type":17,"tag":71,"props":274,"children":276},{"alt":73,"src":275},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110060008.50279295145793742950858514982213:50541109072105:2400:82AC0101EB41149176EEBA506E1EE7B031F1B6EC1CDFD62562031277D3BB2CAE.png",[],{"type":17,"tag":25,"props":278,"children":279},{},[280,285],{"type":17,"tag":31,"props":281,"children":282},{},[283],{"type":23,"value":284},"杨奕博士",{"type":23,"value":286},"，深圳湾实验室副研究员。2010年本科毕业于山东大学。2015年获北京大学理学博士学位，导师为高毅勤教授。2016年1月至2019年2月于瑞士苏黎世联邦理工学院（ETH Zürich）从事博士后研究工作，合作导师为Michele Parrinello教授。2019年4月加入深圳湾实验室系统与物理生物学研究所任副研究员。2020年获华为首批HAE （HUAWEI Ascend Expert）称号，现为昇思MindSpore技术委员会成员、昇思MindSpore资深布道师。",{"type":17,"tag":25,"props":288,"children":289},{},[290,292],{"type":23,"value":291},"主页：",{"type":17,"tag":263,"props":293,"children":296},{"href":294,"rel":295},"https://gitee.com/helloyesterday",[267],[297],{"type":23,"value":294},{"type":17,"tag":25,"props":299,"children":300},{},[301],{"type":17,"tag":31,"props":302,"children":303},{},[304],{"type":23,"value":305},"*本文已被选为”Editor's Choice\",可免费阅读",{"type":17,"tag":25,"props":307,"children":308},{},[309],{"type":17,"tag":31,"props":310,"children":311},{},[312],{"type":23,"value":313},"扫码阅读英文原文",{"type":17,"tag":25,"props":315,"children":316},{},[317],{"type":17,"tag":31,"props":318,"children":319},{},[320],{"type":17,"tag":71,"props":321,"children":323},{"alt":73,"src":322},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110060109.12012807771562901077626117417872:50541109072105:2400:5E0CC995EC834E1EC02DFAA9344421AA2D511E18B257CAEC1ED67B135ABEE35C.png",[],{"type":17,"tag":25,"props":325,"children":326},{},[327,332],{"type":17,"tag":130,"props":328,"children":329},{},[330],{"type":23,"value":331},"J. Chem. Theory Comput.",{"type":23,"value":333}," 2023, 19, 14, 4338–4350",{"type":17,"tag":25,"props":335,"children":336},{},[337],{"type":23,"value":338},"Publication Date: June 26, 2023",{"type":17,"tag":25,"props":340,"children":341},{},[342],{"type":17,"tag":263,"props":343,"children":346},{"href":344,"rel":345},"https://doi.org/10.1021/acs.jctc.3c00214",[267],[347],{"type":23,"value":344},{"type":17,"tag":25,"props":349,"children":350},{},[351],{"type":23,"value":352},"Copyright © 2023 American Chemical Society",{"type":17,"tag":25,"props":354,"children":355},{},[356],{"type":17,"tag":31,"props":357,"children":358},{},[359],{"type":23,"value":360},"关于",{"type":17,"tag":25,"props":362,"children":363},{},[364],{"type":17,"tag":31,"props":365,"children":366},{},[367],{"type":17,"tag":130,"props":368,"children":369},{},[370],{"type":23,"value":220},{"type":17,"tag":25,"props":372,"children":373},{},[374],{"type":17,"tag":31,"props":375,"children":376},{},[377],{"type":17,"tag":130,"props":378,"children":379},{},[380],{"type":17,"tag":71,"props":381,"children":383},{"alt":73,"src":382},"https://fileserver.developer.huaweicloud.com/FileServer/getFile/cmtybbs/e64/154/b38/90a1d5d431e64154b387b3660e356ff5.20231110060155.01061833760376860785498662705211:50541109072105:2400:C6535E9347991D6A0BC2C916CCB6320A3685AC937054877A9F97F8CBEC50B2B0.png",[],{"type":17,"tag":25,"props":385,"children":386},{},[387],{"type":17,"tag":31,"props":388,"children":389},{},[390],{"type":23,"value":391},"影响因子：5.5",{"type":17,"tag":25,"props":393,"children":394},{},[395],{"type":17,"tag":31,"props":396,"children":397},{},[398],{"type":23,"value":399},"主编：",{"type":17,"tag":25,"props":401,"children":402},{},[403],{"type":23,"value":404},"Laura Gagliardi",{"type":17,"tag":25,"props":406,"children":407},{},[408],{"type":23,"value":409},"(University of Chicago)",{"type":17,"tag":25,"props":411,"children":412},{},[413,418],{"type":17,"tag":31,"props":414,"children":415},{},[416],{"type":23,"value":417},"投稿到初审决定平均时间:",{"type":23,"value":419}," 6.4天",{"type":17,"tag":25,"props":421,"children":422},{},[423,431],{"type":17,"tag":31,"props":424,"children":425},{},[426],{"type":17,"tag":130,"props":427,"children":428},{},[429],{"type":23,"value":430},"The Journal of Chemical Theory and Computation",{"type":23,"value":432}," publishes papers reporting new theories, methodology in quantum electronic structure, molecular dynamics, and statistical mechanics and/or their important applications.",{"type":17,"tag":25,"props":434,"children":435},{},[436,438],{"type":23,"value":437},"期刊主页：",{"type":17,"tag":263,"props":439,"children":442},{"href":440,"rel":441},"https://pubs.acs.org/journal/jctcce",[267],[443],{"type":23,"value":440},{"title":7,"searchDepth":445,"depth":445,"links":446},4,[],"markdown","content:news:zh:2854.md","content","news/zh/2854.md","news/zh/2854","md",1776506074119]