2026-09-19T21:57:59Zhttps://dspace.mit.edu/server/oai/request

oai:dspace.mit.edu:1721.1/1197552026-06-06T00:49:30Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023

Bonnie Berger. Chung, Ryan Kyong-doc Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science 2018-12-18T19:48:40Z 2018-12-18T19:48:40Z 2018 2018 http://hdl.handle.net/1721.1/119755 1078691767 Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2018. This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. Cataloged from student-submitted PDF version of thesis. Includes bibliographical references (pages 39-41). Understanding the diversity and abundance of microbial populations is paramount to the health of humans and the environment. Estimating the diversity of these populations from whole metagenome shotgun (WMS) sequencing reads is difficult because the size of these datasets and overlapping reads limit what kinds of analysis we can do. Current methods require matching reads to a database of known microbes. These methods are either too slow or lack the sensitivity needed to identify novel species. We propose a convolutional neural network (CNN) based approach to metagenomic binning that embeds reads into a low-dimensional vector space based on taxonomic classification. We show that our method can get the speed and sensitivity necessary taxonomic classification. Our method was able to achieve 13% accuracy on identifying novel genus of bacteria as compared to 7% accuracy of k-mer embedding. At the same time, the speed of our method is within an order of magnitude of that of k-mer embedding, making it viable as a metagenomic analysis tool. by Ryan Kyong-doc Chung. M.Eng. 41 pages eng Massachusetts Institute of Technology MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission. http://dspace.mit.edu/handle/1721.1/7582 Electrical Engineering and Computer Science. Deep learning approach to metagenomic binning Thesis application/pdf false Publication unknown unknown <Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="4b2aa5af-5f7c-4986-ad82-47153e64e5db"> <Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type> <Language>eng</Language> <Title>Deep learning approach to metagenomic binning</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2018</PublicationDate> <Authors> <Author> <DisplayName>Chung, Ryan Kyong-doc</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <DisplayName>Massachusetts Institute of Technology</DisplayName> <OrgUnit /> </Publisher> </Publishers> <License>http://dspace.mit.edu/handle/1721.1/7582</License> <Keyword>Electrical Engineering and Computer Science.</Keyword> <Abstract>Understanding the diversity and abundance of microbial populations is paramount to the health of humans and the environment. Estimating the diversity of these populations from whole metagenome shotgun (WMS) sequencing reads is difficult because the size of these datasets and overlapping reads limit what kinds of analysis we can do. Current methods require matching reads to a database of known microbes. These methods are either too slow or lack the sensitivity needed to identify novel species. We propose a convolutional neural network (CNN) based approach to metagenomic binning that embeds reads into a low-dimensional vector space based on taxonomic classification. We show that our method can get the speed and sensitivity necessary taxonomic classification. Our method was able to achieve 13% accuracy on identifying novel genus of bacteria as compared to 7% accuracy of k-mer embedding. At the same time, the speed of our method is within an order of magnitude of that of k-mer embedding, making it viable as a metagenomic analysis tool.</Abstract> <Access xmlns="http://purl.org/coar/access_right" > </Access> </Publication>