VirHunter: detection of novel RNA Viruses in Plant Sequencing Data
A Deep Learning-Based Method for the detection of novel RNA Viruses in Plant Sequencing Data
Please welcome VirHunter and Decontaminator as new members on the Galaxy bookshelf. These two tools were developed to streamline the detection of viruses in high-throuput sequencing data (HTS). In particular, they "catch" viruses in assembled virome RNAseq data provided as input.
VirHunter’s role is to classify any contig into one of the three classes: virus, bacteria or plant. It has 7 plant host-specific models and a generalistic model that you can apply to any plant dataset. Decontaminator helps VirHunter as a preprocessing step by eliminating fungi or phage contamination, so they work even better together.
Last but not the least, if you are missing a model in VirHunter and Decontaminator please get in contact with us and we will help you creating new models and add them to Galaxy.
To know more about the tools please visit VirHunter or Decontaminator GitHub repositories. Both tools leverage deep learning for the classification task. We came up with many unique solutions to challenging genomic problems, so if you develop a DL tool for genomics feel free to visit our repositories for inspiration. You can thus use any of the repositories to quickstart your DL project.
Read more about our work in our paper "A Deep Learning-Based Method for Detection of Novel RNA Viruses in Plant Sequencing Data".