MetaMIS

MetaMIS infers microbial interaction networks from next-generation sequencing (NGS) metagenomic time-series operational taxonomic unit (OTU) abundance data to analyze microbial community dynamics.

Key Features:

• Lotka-Volterra Model Integration: MetaMIS employs the Lotka-Volterra model to infer interaction coefficients from OTU abundance tables.
• Handling Missing Data: MetaMIS is robust to high levels of missing data, preserving inference when rare microbes are underrepresented or absent.
• Bray-Curtis Dissimilarity Method: MetaMIS uses Bray-Curtis dissimilarity to evaluate similarity between inferred interaction patterns and biological reference or between networks.
• Interaction Pattern Examination: MetaMIS systematically examines interaction patterns such as mutualism and competition to characterize biotic roles within communities.
• Consensus Network Formation: MetaMIS aggregates multiple interaction networks into a consensus network to facilitate comparative studies across samples or conditions, as demonstrated in comparisons of female and male fecal microbiomes.

Scientific Applications:

• Microbial ecology research: MetaMIS reveals species interactions and community dynamics from metagenomic time-series data.
• Time-series interaction discovery: MetaMIS identifies critical interactions that may be missed by static analyses using OTU time-series.
• Human fecal microbiome studies: MetaMIS has been applied to compare female and male fecal microbiomes and identified Micrococcaceae as a potentially important low-abundance player.

Methodology:

MetaMIS infers interactions from OTU abundance data using the Lotka-Volterra model, interprets inferred interactions within a network framework, employs Bray-Curtis dissimilarity to assess biological relevance, and incorporates procedures to handle high levels of missing data.