Continuous Embeddings of DNA Sequencing Reads and Application to Metagenomics

Journal of computational biology: A journal of computational molecular cell biology

Jean-PhilippeVert^3,1,2,4,5⁴  RomainMenegaux^1,2⁵ 

[1] Address correspondence to: Dr. Jean–Philippe Vert, Google Brain, 8 rue de Londres, 75009 Paris, France^3;Ecole Normale Supérieure, Department of Mathematics and Applications, CNRS, PSL Research University, Paris, France^4;Google Brain, Paris, France^5;Institut Curie, PSL Research University, INSERM, U900, Paris, France^2;MINES ParisTech, PSL Research University, CBIO—Centre for Computational Biology, Paris, France^1

关键词: metagenomics;    sequencing;    classification;    embedding.;   

DOI  :  10.1089/cmb.2018.0174

学科分类：生物科学（综合）

来源: Mary Ann Liebert, Inc. Publishers

【 摘 要 】

We propose a new model for fast classification of DNA sequences output by next-generation sequencing machines. The model, which we call fastDNA, embeds DNA sequences in a vector space by learning continuous low-dimensional representations of thek-mers it contains. We show on metagenomics benchmarks that it outperforms the state-of-the-art methods in terms of accuracy and scalability.

【 授权许可】

Unknown