Cloud-Based Solution for Improving Usability and Interactivity of Metagenomic Ensemble Taxonomic Classification Methods

Asem Abdelaziz1, Ahmed Metwally2

  • 1Cairo University, Faculty of Engineering
  • 2University of Illinois at Chicago

Details

12:10 - 12:25 | Tue 6 Mar | Treasure Island ABC | TuAT1.5

Session: BHI Session # 3 – Bioinformatics for Health Monitoring

Abstract

The microbiome plays a vital role in a broad range of host-related processes and has a significant effect on host health. Many algorithms have been developed to classify the origin of the microbial sequences and construct the sample's microbial profile. Since these methods individually are suboptimal and their classification results are diverse, more attempts are devoted to develop ensemble methods for enhancing classification quality. The research associated with developing ensemble methods is mainly targeting multiple performance criteria, such as precision and sensitivity. However, none of these methods have presented a usability-oriented design targeting most biologists who lack adequate computational skills. In this paper, we present cloud-based framework of ensemble taxonomic classification method. In addition, the implementation presents a platform for interactive visualization of the generated results. The project repository is publicly available on https://github.com/aametwally/WEVOTE-web.