metaVaR

metaVaR implements the metavariant species (MVS) model to perform reference-free population genomic analyses of uncultured organisms directly from multisample metagenomic raw reads.

Key Features:

• Implementation: Implemented as an R package that encapsulates the MVS model for population genomics from metagenomic data.
• Reference-Free Analysis: Analyzes intra-species nucleotide polymorphisms without relying on reference genomes or assembled reads.
• Metavariant Species (MVS) Model: Represents species via their unique intra-species variants clustered from multisample metagenomic raw reads.
• Clustering Algorithms: Combines reference-free variant calling with multiple density-based clustering techniques and maximum weighted independent set algorithms to group intra-species variants into MVSs.
• Population Genomic Statistics: Uses MVS frequencies to compute statistics such as FST for estimating genomic differentiation and identifying loci under natural selection.
• Robustness and Accuracy: Validated on simulated and real metagenomic datasets with reported ΔFST < 0.0001 on simulated data and ΔFST < 0.03 on real datasets, and with loci predictions validated against real data.

Scientific Applications:

• Population genomics of uncultured organisms: Enables population-level analyses of microorganisms that lack reference genomes or transcriptomes.
• Study of small eukaryotes: Facilitates genomic investigations of small eukaryotes important to ecological systems but difficult to culture or sequence conventionally.
• Detection of genomic differentiation and selection: Supports estimation of genomic differentiation (FST) across populations and identification of loci putatively under natural selection.

Methodology:

Focuses on intra-species nucleotide polymorphisms; performs reference-free variant calling; clusters variants from multisample metagenomic raw reads using density-based clustering and maximum weighted independent set algorithms to construct MVSs; estimates MVS frequencies to compute FST and identify loci under selection.