Performance, Accuracy, and Web Server for Evolutionary Placement of Short Sequence Reads under Maximum Likelihood

Abstract

We present an evolutionary placement algorithm (EPA) and a Web server for the rapid assignment of sequence fragments (short reads) to edges of a given phylogenetic tree under the maximum-likelihood model. The accuracy of the algorithm is evaluated on several real-world data sets and compared with placement by pair-wise sequence comparison, using edit distances and BLAST. We introduce a slow and accurate as well as a fast and less accurate placement algorithm. For the slow algorithm, we develop additional heuristic techniques that yield almost the same run times as the fast version with only a small loss of accuracy. When those additional heuristics are employed, the run time of the more accurate algorithm is comparable with that of a simple BLAST search for data sets with a high number of short query sequences. Moreover, the accuracy of the EPA is significantly higher, in particular when the sample of taxa in the reference topology is sparse or inadequate. Our algorithm, which has been integrated into RAxML, therefore provides an equally fast but more accurate alternative to BLAST for tree-based inference of the evolutionary origin and composition of short sequence reads. We are also actively developing a Web server that offers a freely available service for computing read placements on trees using the EPA.

Parasites dominate hyperdiverse soil protist communities in Neotropical rainforests

“…Furthermore, PR 2 and similar databases are biased towards marine and temperate references. To solve these problems, reads were dereplicated into 10.6 million amplicons (strictly identical reads to which an abundance value can be attached), and subsequently placed using the Evolutionary Placement Algorithm (EPA) 14 , as implemented in RAxML 15 , onto a phylogenetic tree inferred from 512 full-length references from all major eukaryotic clades ( Fig. 2 and Supplementary Fig.…”

“…In brief, we used RAxML 15 v8.1.15 to infer reference trees from the reference alignments, then used PaPaRa 70 v2.4 to align the query sequences to those reference alignments, and finally used the EPA 14 as implemented in RAxML to place the queries onto the trees. Placement results were visualized as 'heat trees' using Genesis 71 v0.2.0.…”

Microbial community successional patterns in beach sands impacted by the Deepwater Horizon oil spill

“…The resulting dataset for AlkB was aligned using Muscle v3.8.31 with default parameters (Edgar, 2004), and the gene phylogeny was reconstructed using RAxML v7.7.2 with GTR model (proteins), gamma parameter optimization, and '-f a' algorithm (Stamatakis, 2006). Putative coding fragments predicted with FragGeneScan (Rho et al, 2010) on sequence reads were subsequently placed onto the reconstructed tree based on a sequence-to-profile alignment built with Clustal Omega v1.1.0 (Sievers et al, 2011), using the evolutionary placement algorithm (Berger et al, 2011). The same placement strategy was independently applied to the partial sequences of AlkB reported in the study by Smith et al (2013) (GenBank entries KF613175-KF613575).…”

Re-evaluating the phylogeny of Sipuncula through transcriptomics

“…To test for putative gene incongruence, we inferred individual gene trees for each ortholog group included in each of the two matrices using RAxML 7.7.5 (Berger et al, 2011) and the PROTGAMMALG4X model of selection. All individual best-scoring trees were concatenated per matrix (one file for the 50% taxon-occupancy matrix containing 675 genes, and one for the 75% taxon-occupancy matrix containing 141 genes) and fed into SuperQ v1.1 (Grünewald et al, 2013) in order to visualize intergene conflict.…”