Compressive pangenomics using mutation-annotated networks.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Pangenomics aims to capture intra-species diversity using genome collections, moving beyond single references.
• Current pangenomic formats struggle with scalability to millions of genomes and often overlook mutation details and evolutionary links.

Purpose of the Study:

• Introduce Pangenome Mutation-Annotated Network (PanMAN) as a novel, lossless pangenomic data representation.
• Enhance the capacity of pangenomic formats to encode mutational and evolutionary histories.
• Enable scalable and efficient pangenomic analysis for large datasets.

Main Methods:

• Developed PanMAN, a lossless pangenome representation encoding mutational and evolutionary histories.
• Achieved significant file size compression compared to existing variation-preserving formats.
• Constructed a comprehensive SARS-CoV-2 pangenome from 8 million sequences using PanMAN.
• Introduced 'panmanUtils' toolkit for analysis and interoperability.

Main Results:

• PanMAN achieved file size compression ratios from 3.5-1,391×.
• Compression performance generally improved with larger datasets.
• A SARS-CoV-2 pangenome of 8 million sequences required only 366 MB using PanMAN.
• PanMAN enables detailed encoding of mutational and evolutionary histories.

Conclusions:

• PanMAN significantly improves the scale, speed, resolution, and scope of pangenomic analysis.
• The format facilitates efficient data sharing and biological discovery.
• PanMAN and its associated toolkit are poised to advance the field of pangenomics.