Deep6: Classification of Metatranscriptomic Sequences into Cellular Empires and Viral Realms Using Deep Learning Models

Abstract

Metatranscriptomic data is increasingly used to study viral diversity and activity; yet, identifying and taxonomically assigning viral sequences is still challenging. Deep6 is a deep-learning model that classifies metatranscriptomic sequences into six groups: prokaryotes, eukaryotes, or one of the four viral realms. Deep6 is trained on reference coding sequences, but classification of query sequences is done reference-independent and alignment-free. The provided model is optimized for marine samples and can process sequences as short as 250 nucleotides. Average accuracies range from 0.87 to 0.97 depending on sequence length. Additionally, Deep6 includes scripts to easily encode and train custom models for other environments.