He Hao, Liu Can, Liu Haiguang
Complex Systems Division, Beijing Computational Science Research Center, 8 E Xibeiwang Road, Haidian, Beijing 100193, People's Republic of China; School of Software Engineering, University of Science and Technology China, Suzhou, Jiang Su 215123, People's Republic of China.
Complex Systems Division, Beijing Computational Science Research Center, 8 E Xibeiwang Road, Haidian, Beijing 100193, People's Republic of China; Physics Department, Beijing Normal University, Haidian, Beijing 100875, People's Republic of China.
iScience. 2020 Mar 27;23(3):100906. doi: 10.1016/j.isci.2020.100906. Epub 2020 Feb 13.
Small-angle X-ray scattering (SAXS) method is widely used in investigating protein structures in solution, but high-quality 3D model reconstructions are challenging. We present a new algorithm based on a deep learning method for model reconstruction from SAXS data. An auto-encoder for protein 3D models was trained to compress 3D shape information into vectors of a 200-dimensional latent space, and the vectors are optimized using genetic algorithms to build 3D models that are consistent with the scattering data. The program has been tested with experimental SAXS data, demonstrating the capacity and robustness of accurate model reconstruction. Furthermore, the model size information can be optimized using this algorithm, enhancing the automation in model reconstruction directly from SAXS data. The program was implemented using Python with the TensorFlow framework, with source code and webserver available from http://liulab.csrc.ac.cn/decodeSAXS.
小角X射线散射（SAXS）方法被广泛用于研究溶液中的蛋白质结构，但高质量的三维模型重建具有挑战性。我们提出了一种基于深度学习方法的新算法，用于从SAXS数据进行模型重建。训练了一个用于蛋白质三维模型的自动编码器，将三维形状信息压缩到一个200维潜在空间的向量中，并使用遗传算法对这些向量进行优化，以构建与散射数据一致的三维模型。该程序已用实验SAXS数据进行了测试，证明了准确模型重建的能力和稳健性。此外，使用该算法可以优化模型大小信息，直接从SAXS数据增强模型重建的自动化。该程序使用Python和TensorFlow框架实现，源代码和网络服务器可从http://liulab.csrc.ac.cn/decodeSAXS获得。
