Phylen: automatic phylogenetic reconstruction using the EggNOG database

Abstract

The EggNOG database (Powell et al. 2014) is an excellent resource providing orthologous groups shared at different taxonomic ranks including several prokaryotes. Here we present Phylen, a simple and automated software package written in R that reconstructs phylogenies by interacting with the EggNOG database. First, a set of orthologous groups available at the EggNOG database is selected and automatically downloaded or, alternatively, an external set of orthologous groups can be provided formatted as a Hidden Markov Model (HMM) file. Second, genome annotations in GFF3 format (such as those from Prokka annotation software (Seemann 2014)) are parsed to extract translated coding sequences. Third, genomes are screened against these orthologous groups using HMMER3 (Eddy 2011). Forth, “core” coding sequences are extracted and multiple sequence alignment is performed over each recovered gene set using MAFFT (Katoh and Standley 2013). Fifth, alignments are concatenated into a single supergene and phylogenetic reconstruction is performed using Maximum-Likelihood or distance methods (Fig. 1A). Phylen outputs one multi-fasta alignment per gene, one supergene multi-fasta alignment file, one tree file in Newick format and an object of class “phylo” which can be further analysed using the R packeges ape (Paradis, Claude, and Strimmer 2004) and phangorn (Schliep 2011).