Identification of mobile genetic elements with geNomad

Area of Science:

• Genomics
• Bioinformatics
• Virology

Background:

• Mobile genetic elements (MGEs) like plasmids and viruses are crucial for evolution, ecology, and public health.
• Accurate identification and characterization of MGEs from sequencing data are vital but challenging.
• Existing tools often lack the precision and scalability needed for large-scale genomic analyses.

Purpose of the Study:

• To introduce geNomad, a novel computational framework for classifying and annotating plasmids and viruses.
• To enhance the detection of MGEs, including integrated proviruses, within host genomes.
• To provide a scalable and high-performance solution for analyzing massive sequencing datasets.

Main Methods:

• geNomad integrates gene content analysis with a deep neural network for MGE identification.
• It utilizes a comprehensive dataset of over 200,000 marker protein profiles for functional annotation and taxonomic assignment.
• A conditional random field model is employed for precise detection of integrated proviruses.

Main Results:

• geNomad demonstrated superior performance in classifying plasmids (MCC 77.8%) and viruses (MCC 95.3%) compared to other tools.
• The framework successfully processed over 2.7 trillion base pairs of sequencing data.
• This analysis led to the discovery of millions of novel viruses and plasmids.

Conclusions:

• geNomad offers a highly accurate, fast, and scalable method for identifying and annotating MGEs in large sequencing datasets.
• The discovered viral and plasmid sequences are publicly available via the IMG/VR and IMG/PR databases.
• geNomad represents a significant advancement in the study of MGEs and their impact.