2020/06/15 by Amélie Chatelain, Chatelain, Amélie, Tommasone, Elena +5
Computer Science · Biochemistry, Genetics and Molecular Biology · #Computational Drug Discovery Methods #Metabolomics and Mass Spectrometry Studies #Gene Regulatory Network Analysis
paper · pdf · doi:10.48550/arxiv.2006.08697
Proteins are made of atoms constantly fluctuating, but can occasionally\nundergo large-scale changes. Such transitions are of biological interest,\nlinking the structure of a protein to its function with a cell. Atomic-level\nsimulations, such as Molecular Dynamics (MD), are used to study these events.\nHowever, molecular dynamics simulations produce time series with multiple\nobservables, while changes often only affect a few of them. Therefore,\ndetecting conformational changes has proven to be challenging for most\nchange-point detection algorithms. In this work, we focus on the identification\nof such events given many noisy observables. In particular, we show that the\nNo-prior-Knowledge Exponential Weighted Moving Average (NEWMA) algorithm can be\nused along optical hardware to successfully identify these changes in\nreal-time. Our method does not need to distinguish between the background of a\nprotein and the protein itself. For larger simulations, it is faster than using\ntraditional silicon hardware and has a lower memory footprint. This technique\nmay enhance the sampling of the conformational space of molecules. It may also\nbe used to detect change-points in other sequential data with a large number of\nfeatures.\n