Abstract |
The animal gut is considered to be one of the most densely populated by microorganisms
ecosystem. All this microbial weight, vital for host's health and homeostasis, has been co-evolving
along with the gut structure, for millions years now. Recent studies on human gut microbiome have
already demonstrated the genetic repertoire of this community and highlighted its genetic variation
among individuals, not only in health states but also in various disease conditions. However, the
evolutionary processes that drive the genetic diversity of the gut microbiome have not been
thoroughly investigated. Here, we use metagenomic datasets from stool samples of 39 healthy US
individuals to construct the genomic variation landscape of the most abundant bacterial species in
the human gut. Our main goals are, the detection of events of strong positive selection, based on the
site frequency spectrum (SFS) and the inference of a demographic model for each one of the
bacterial species comprising the metagenomic sample. To infer the demographic scenarios we use
an approximate computation bayesian technique based on coalescent simulations, taking into
account the SFS. We find that the abundance or even presence of previously reported as highly
abundant gut species varies among the individuals in our cohort. We report the most possible
demographic scenarios for each bacterial strains, which, for most of the bacteria, comprises a
recent, sharp population size decline. Furthermore, we report our findings on the detection of strong
positive selection. Finally,, we report a plethora of difficulties in the estimation of SFS from
metagenomic samples, such as the method for estimating allele frequencies from datasets with
missing observations.
|