Detecting selection in bacteriophage genomes using SNPs from time-series metagenomic data
Metagenomes, the genetic content from community members in a specific environment, contain unprecedented information on the genetic variation within community members. Studying genome evolution in unculturable organisms, including bacteria and their bacteriophages, requires methods and models for the analysis of metagenomes using evolutionary and population genetics theory.
We present a novel method to detect directional selection in time-series metagenomic data. Single-nucleotide polymorphisms (SNPs) are extracted from the metagenomic read information and their frequency at each time point is calculated. The rate of SNP frequency change over time is estimated from the linear regression of the SNP frequency as a function of time. Simulations show that random genetic drift results in slope distributions with mean zero that can be approximated by normal distributions. In contrast, in the presence of selection, the slope distributions show a mean different from zero.
Our approach is applied to phage contigs identified by single-cell sequencing (Roux et al. eLife 2014;3:e03125). Utilizing available time-series metagenomic data, we reconstruct the evolution and population dynamics of the phage sequences. We analyse two phage contigs present in high coverage in the metagenomes. Genetic heterogeneity is abundant within these contigs with 10% and 13% of the positions exhibiting polymorphisms, respectively. The evaluation of models excluding and including selection suggests that one phage contig evolved under selection over the sampling duration. In this contig, breakpoints can be identified that localize selection to a region encoding the phage structural proteins. Random genetic drift explains the evolutionary trajectory of the other contig.
We show that adaptation dynamics occurring during the course of time-series metagenomic samplings can be detected by the analysis of SNP frequencies. Different phages in the same environment are exposed to different selective pressures. Our approach opens new possibilities for the research of bacteria and phage evolution from metagenomics data.
We present a novel method to detect directional selection in time-series metagenomic data. Single-nucleotide polymorphisms (SNPs) are extracted from the metagenomic read information and their frequency at each time point is calculated. The rate of SNP frequency change over time is estimated from the linear regression of the SNP frequency as a function of time. Simulations show that random genetic drift results in slope distributions with mean zero that can be approximated by normal distributions. In contrast, in the presence of selection, the slope distributions show a mean different from zero.
Our approach is applied to phage contigs identified by single-cell sequencing (Roux et al. eLife 2014;3:e03125). Utilizing available time-series metagenomic data, we reconstruct the evolution and population dynamics of the phage sequences. We analyse two phage contigs present in high coverage in the metagenomes. Genetic heterogeneity is abundant within these contigs with 10% and 13% of the positions exhibiting polymorphisms, respectively. The evaluation of models excluding and including selection suggests that one phage contig evolved under selection over the sampling duration. In this contig, breakpoints can be identified that localize selection to a region encoding the phage structural proteins. Random genetic drift explains the evolutionary trajectory of the other contig.
We show that adaptation dynamics occurring during the course of time-series metagenomic samplings can be detected by the analysis of SNP frequencies. Different phages in the same environment are exposed to different selective pressures. Our approach opens new possibilities for the research of bacteria and phage evolution from metagenomics data.
Reference:
Molecular Interactions-T06-Oft-03
Session:
Molecular Interactions between viruses and their host cells
Presenters:
Anne Kupczok
Session:
Molecular interactions between viruses and their host cells
Presentation type:
Offered talk - 15 min
Room:
Main Auditorium
Chair/s:
Heather Allison
Date:
Wednesday, 20 July 2016
Time:
09:25 - 09:40