BioAutoMATED: An end-to-end automated machine learning tool for explanation and design of biological sequences

作     者：Valeri, Jacqueline A. Soenksen, Luis R. Collins, Katherine M. Ramesh, Pradeep Cai, George Powers, Rani Angenent-Mari, Nicolaas M. Camacho, Diogo M. Wong, Felix Lu, Timothy K. Collins, James J. 

作者机构：MIT Dept Biol Engn 77 Massachusetts Ave Cambridge MA 02139 USA MIT Inst Med Engn & Sci 77 Massachusetts Ave Cambridge MA 02139 USA Harvard Univ Wyss Inst Biolog Inspired Engn Boston MA 02115 USA Broad Inst & Harvard Cambridge MA 02142 USA MIT Dept Mech Engn 77 Massachusetts Ave Cambridge MA 02139 USA MIT Dept Elect Engn & Comp Sci 77 Massachusetts Ave Cambridge MA 02139 USA MIT Dept Brain & Cognit Sci 77 Massachusetts Ave Cambridge MA 02139 USA Univ Cambridge Dept Engn Trumpington St Cambridge CB2 1PZ England Pluto Biosci Golden CO 80402 USA MIT Res Lab Elect Synthet Biol Grp Cambridge MA 02139 USA Harvard MIT Program Hlth Sci & Technol Cambridge MA 02139 USA MIT Abdul Latif Jameel Clin Machine Learning Hlth Cambridge MA 02139 USA 

出 版 物：《CELL SYSTEMS》 (Cell Syst.)

年 卷 期：2023年第14卷第6期

页      面：525-+页

核心收录：

学科分类：0710[理学-生物学] 07[理学] 071009[理学-细胞生物学] 09[农学] 0901[农学-作物学] 090102[农学-作物遗传育种] 

基　　金：Defense Threat Reduction Agency [HDTRA-12210032] DARPA SD2 program Paul G. Allen Frontiers Group Wyss Institute for Biologically Inspired Engineering, Harvard University Audacious Project Flu Lab, LLC Sea Grape Foundation MIT-Takeda Fellowship Siebel Foundation Scholarship CONACyT [342369/408970] MIT Tata Center fellowship Barry Goldwater Scholarship Marshall Scholarship Cambridge Trust National Institute of Allergy and Infectious Diseases of the National Institutes of Health [K25AI168451] 

主　　题：automated machine learning architecture search hyperparameter optimization biological sequences 

摘      要：The design choices underlying machine-learning (ML) models present important barriers to entry for many biologists who aim to incorporate ML in their research. Automated machine-learning (AutoML) algorithms can address many challenges that come with applying ML to the life sciences. However, these algorithms are rarely used in systems and synthetic biology studies because they typically do not explicitly handle biological sequences (e.g., nucleotide, amino acid, or glycan sequences) and cannot be easily compared with other AutoML algorithms. Here, we present BioAutoMATED, an AutoML platform for biological sequence analysis that integrates multiple AutoML methods into a unified framework. Users are automat-ically provided with relevant techniques for analyzing, interpreting, and designing biological sequences. BioAutoMATED predicts gene regulation, peptide-drug interactions, and glycan annotation, and designs optimized synthetic biology components, revealing salient sequence characteristics. By auto-mating sequence modeling, BioAutoMATED allows life scientists to incorporate ML more readily into their work.