Phylogeny-Aware Deep 1-Dimensional Convolutional Neural Network for the Classification of Metagenomes

Timmy Manning, Jyotsna Talreja Wassan, Cintia Palu, Haiying / HY Wang, Browne Fiona, Huiru Zheng, Brian Kelly, Paul Walsh

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper evaluates a novel approach to the integration of biological domain knowledge relating to the natural evolutionary structure of microbial community data to classifying 16S rDNA sequence samples. Specifically, we evaluate the use of phylogenetic trees in addition to amplicon sequence variant abundance in samples for the classification of a processed cattle metagenomics data set using machine learning. Further to this, we employ a class activation map of the network when applied to specific exemplars to determine, firstly, the relevance of higher level taxonomic data, and secondly, the most relevant taxa in determining the classification, according to the classifier.
Original languageEnglish
Title of host publicationProceedings of 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)
PublisherIEEE
Pages1826-1831
Number of pages6
ISBN (Electronic)978-1-5386-5488-0, 978-1-5386-5487-3
ISBN (Print)978-1-5386-5489-7
DOIs
Publication statusPublished - 3 Dec 2018
EventIEEE International Conference on Bioinformatics and Biomedicine (BIBM) - Madrid, Spain
Duration: 3 Dec 20186 Dec 2018
http://orienta.ugr.es/bibm2018/

Conference

ConferenceIEEE International Conference on Bioinformatics and Biomedicine (BIBM)
CountrySpain
CityMadrid
Period3/12/186/12/18
Internet address

    Fingerprint

Keywords

  • Phylogeny
  • Feature extraction
  • Bioinformatics
  • Microorganisms
  • Image coding
  • Genomics ,
  • Machine learning

Cite this

Manning, T., Wassan, J. T., Palu, C., Wang, H. . HY., Fiona, B., Zheng, H., ... Walsh, P. (2018). Phylogeny-Aware Deep 1-Dimensional Convolutional Neural Network for the Classification of Metagenomes. In Proceedings of 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (pp. 1826-1831). IEEE. https://doi.org/10.1109/BIBM.2018.8621543