PhyloSift: phylogenetic analysis of genomes and metagenomes

Abstract

Like all organisms on the planet, environmental microbes are subject to the forces of molecular evolution. Metagenomic sequencing provides a means to access the DNA sequence of uncultured microbes. By combining DNA sequencing of microbial communities with evolutionary modeling and phylogenetic analysis we might obtain new insights into microbiology and also provide a basis for practical tools such as forensic pathogen detection. In this work we present an approach to leverage phylogenetic analysis of metagenomic sequence data to conduct several types of analysis. First, we present a method to conduct phylogeny-driven Bayesian hypothesis tests for the presence of an organism in a sample. Second, we present a means to compare community structure across a collection of many samples and develop direct associations between the abundance of certain organisms and sample metadata. Third, we apply new tools to analyze the phylogenetic diversity of microbial communities and again demonstrate how this can be associated to sample metadata. These analyses are implemented in an open source software pipeline called PhyloSift. As a pipeline, PhyloSift incorporates several other programs including LAST, HMMER, and pplacer to automate phylogenetic analysis of protein coding and RNA sequences in metagenomic datasets generated by modern sequencing platforms (e.g., Illumina, 454).

Keywords: Bayes factor; Community structure; Edge PCA; Forensics; Metagenomics; Microbial diversity; Microbial ecology; Microbial evolution; Phylogenetic diversity; Phylogenetics.

Figure 1. PhyloSift client workflow.

This workflow is applied to the user’s sequence data. DNA input sequences are processed via both the rRNA and protein parts of the workflow.

Figure 2. Comparison of QIIME PCA and edge PCA analysis of human fecal samples.

Samples from 106 individuals were analyzed by PCA to evaluate trends in community composition with respect to host age. 16S rDNA amplicon data and metagenomic data from the same samples was processed using QIIME and PhyloSift. QIIME analyzed the amplicon data (top left) and 16S rDNA reads extracted from the metagenomic data (top right) using a reference-based OTU picking strategy. PhyloSift analyzed the same metagenomic 16S rDNA reads (bottom left) and reads matching the 37 elite gene families (bottom right). Each PCA approach gives qualitatively similar results, differences as quantified by Procrustes analysis are given in Table 1.

Figure 3. Lineages contributing variation in human fecal sample community structure.

106 metagenomic samples were processed using PhyloSift and their community composition compared using Edge PCA (Matsen & Evans, 2013). Lineages that decrease in abundance along the principal component axis are shown in turquoise color, those increasing in abundance are shown in red. Edge width is proportional to the change in abundance. Remaining lineages in the phylogeny of bacteria, archaea, eukarya, and some viruses are shown in light gray. PC1 shown at left, PC2 at right.

Figure 4. Relationship between fecal community phylogenetic diversity and host age.

106 metagenomic samples were processed using PhyloSift and their phylogenetic diversity analyzed using two metrics. Unweighted phylogenetic diversity (PD) simply measures the total branch length of the reference tree covered by placed reads from a sample. Balance-weighted phylogenetic diversity adjusts these values by the abundance of each lineage in the sample. In unweighted PD, a log-linear relationship between host age and fecal community phylogenetic diversity can be observed. Balance weighted PD, on the other hand, shows rapid growth in early life followed by slow decline after the first year, consistent with a small number of divergent lineages becoming dominant in the fecal ecosystem.

Figure 5. Taxonomic visualization of two human gut samples.

Taxonomic plot at left shows an infant, plot at right shows a 45 year old mother. Data analyzed by PhyloSift, visualized by Krona.

Figure 6. PhyloSift performance and scaling behavior.

PhyloSift v1.0 was used to process Illumina sequence data from a human gut microbiome dataset subsampled to varying numbers of reads. The program was run single-threaded on an Intel Xeon E5520 CPU core (circa 2009 model).