2020/08/11 by Alexander Hudson, D. R. Hudson, Hudson, Alexander +2
Biochemistry, Genetics and Molecular Biology · Computer Science · Engineering · Materials Science · #Algorithm #Artificial intelligence #Biomolecules (q-bio.BM) #Boosting (machine learning) #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Convolutional neural network #Enzyme Structure and Function #FOS: Biological sciences #FOS: Computer and information sciences #Function (biology) #Geography #High resolution #I.2 #I.4 #Image Processing Techniques and Applications #Low resolution #Machine Learning (cs.LG) #Machine learning #Nuclear magnetic resonance #Pattern recognition (psychology) #Physics #Protein Structure and Dynamics #Protein structure #Remote sensing #Representation (politics) #Resolution (logic) #cs.CV #cs.LG #q-bio.BM
paper · pdf · doi:10.48550/arxiv.2008.04757
published in arXiv (Cornell University) (Cornell University) · 9 pages excluding references and appendices
openalex publication_date 2020/08/11 · arxiv created 2020/08/31 · arxiv updated 2020/09/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Structure determination is key to understanding protein function at a molecular level. Whilst significant advances have been made in predicting structure and function from amino acid sequence, researchers must still rely on expensive, time-consuming analytical methods to visualise detailed protein conformation. In this study, we demonstrate that it is possible to make accurate (≥80%) predictions of protein class and architecture from structures determined at low (>3A) resolution, using a deep convolutional neural network trained on high-resolution (≤3A) structures represented as 2D matrices. Thus, we provide proof of concept for high-speed, low-cost protein structure classification at low resolution, and a basis for extension to prediction of function. We investigate the impact of the input representation on classification performance, showing that side-chain information may not be necessary for fine-grained structure predictions. Finally, we confirm that high-resolution, low-resolution and NMR-determined structures inhabit a common feature space, and thus provide a theoretical foundation for boosting with single-image super-resolution.