Direct generation of protein conformational ensembles via machine learning.
Direct generation of protein conformational ensembles via machine learning.
复制标题
DOI：
10.1038/s41467-023-36443-x
复制
发表时间：
2023-02-11
影响因子：
16.6
通讯作者：
Feig, Michael
中科院分区：
文献类型：
--
作者：
Janson, Giacomo;Valdes-Garcia, Gilberto;Heo, Lim;Feig, Michael
动力学和构象取样对于将蛋白质结构与生物学功能联系起来是必不可少的。虽然具有挑战性的实验探测，计算机模拟被广泛用于描述蛋白质动力学，但在显着的计算成本，继续限制系统可以研究。在这里，我们证明了机器学习可以用模拟数据进行训练，直接生成物理上真实的蛋白质构象集合，而不需要任何采样，计算成本可以忽略不计。作为原理证明，我们训练了一个基于Transformer架构的生成对抗网络，该网络具有对本质无序肽的粗粒度模拟的自注意力。由此产生的模型，idpGAN，可以预测序列依赖的粗粒度的合奏序列中不存在的训练集证明，可转移性可以实现超出有限的训练数据。我们还重新训练idpGAN原子模拟数据表明，该方法可以扩展到更高分辨率的构象系综生成的原则。研究蛋白质结构动力学的计算方法在生命科学中是一种强有力的工具，但计算成本很高。在这里，作者表明，机器学习可用于有效地生成蛋白质构象集合，并在内在无序的肽上测试他们的方法。
Dynamics and conformational sampling are essential for linking protein structure to biological function. While challenging to probe experimentally, computer simulations are widely used to describe protein dynamics, but at significant computational costs that continue to limit the systems that can be studied. Here, we demonstrate that machine learning can be trained with simulation data to directly generate physically realistic conformational ensembles of proteins without the need for any sampling and at negligible computational cost. As a proof-of-principle we train a generative adversarial network based on a transformer architecture with self-attention on coarse-grained simulations of intrinsically disordered peptides. The resulting model, idpGAN, can predict sequence-dependent coarse-grained ensembles for sequences that are not present in the training set demonstrating that transferability can be achieved beyond the limited training data. We also retrain idpGAN on atomistic simulation data to show that the approach can be extended in principle to higher-resolution conformational ensemble generation. Computational methods to study protein structural dynamics are a powerful tool in life sciences but are computationally expensive. Here, the authors show that machine learning can be used to efficiently generate protein conformational ensembles and test their method on intrinsically disordered peptides.
登录
查看更多内容
影响因子：
64.8
作者：
Jumper J;Evans R;Pritzel A;Green T;Figurnov M;Ronneberger O;Tunyasuvunakool K;Bates R;Žídek A;Potapenko A;Bridgland A;Meyer C;Kohl SAA;Ballard AJ;Cowie A;Romera-Paredes B;Nikolov S;Jain R;Adler J;Back T;Petersen S;Reiman D;Clancy E;Zielinski M;Steinegger M;Pacholska M;Berghammer T;Bodenstein S;Silver D;Vinyals O;Senior AW;Kavukcuoglu K;Kohli P;Hassabis D
通讯作者：
Hassabis D
影响因子：
64.8
作者：
Eisenmesser, EZ;Millet, O;Kern, D
通讯作者：
Kern, D
影响因子：
7.7
作者：
Dutagaci B;Nawrocki G;Goodluck J;Ashkarran AA;Hoogstraten CG;Lapidus LJ;Feig M
通讯作者：
Feig M
影响因子：
4.3
作者：
Eastman P;Swails J;Chodera JD;McGibbon RT;Zhao Y;Beauchamp KA;Wang LP;Simmonett AC;Harrigan MP;Stern CD;Wiewiora RP;Brooks BR;Pande VS
通讯作者：
Pande VS
CHARMM：生物分子模拟程序。
影响因子：
3
作者：
Brooks, B. R.;Brooks, C. L., III;Mackerell, A. D., Jr.;Nilsson, L.;Petrella, R. J.;Roux, B.;Won, Y.;Archontis, G.;Bartels, C.;Boresch, S.;Caflisch, A.;Caves, L.;Cui, Q.;Dinner, A. R.;Feig, M.;Fischer, S.;Gao, J.;Hodoscek, M.;Im, W.;Kuczera, K.;Lazaridis, T.;Ma, J.;Ovchinnikov, V.;Paci, E.;Pastor, R. W.;Post, C. B.;Pu, J. Z.;Schaefer, M.;Tidor, B.;Venable, R. M.;Woodcock, H. L.;Wu, X.;Yang, W.;York, D. M.;Karplus, M.
通讯作者：
Karplus, M.