MPI-blastn and NCBI-TaxCollector: Improving metagenomic analysis with high performance classification and wide taxonomic attachment

Abstract

Metagenomic sequencing technologies are advancing rapidly and the size of output data from high-throughput genetic sequencing has increased substantially over the years. This brings us to a scenario where advanced computational optimizations are requested to perform a metagenomic analysis. In this paper, we describe a new parallel implementation of nucleotide BLAST (MPI-blastn) and a new tool for taxonomic attachment of Basic Local Alignment Search Tool (BLAST) results that supports the NCBI taxonomy (NCBI-TaxCollector). MPI-blastn obtained a high performance when compared to the mpiBLAST and ScalaBLAST. In our best case, MPI-blastn was able to run 408 times faster in 384 cores. Our evaluations demonstrated that NCBI-TaxCollector is able to perform taxonomic attachments 125 times faster and needs 120 times less RAM than the previous TaxCollector. Through our optimizations, a multiple sequence search that currently takes 37 hours can be performed in less than 6 min and a post processing with NCBI taxonomic data attachment, which takes 48 hours, now is able to run in 23 min.

Culture-Independent Analysis of Bacterial Diversity during Bioremediation of Soil Contaminated with a Diesel-Biodiesel Blend (B10)S

“…This tool consists of an adapted version of the algorithm introduced in NCBI-taxcollector [29], which includes an additional step before the taxonomic assignment. In this additional step, all sequence accession numbers from SILVA database were mapped with taxonomic identification numbers (TAXID) from NCBI taxonomy database [30].Sequence matches were classified at an 80% identity level for domain and phylum; 90% identity for class, order, and family; 95% identity for genus; and a 99% identity level for species. The total numbers of 16S rRNA classified sequences were converted into an OTU abundance matrix for each taxonomy level across the samples.…”

Different combinations of atomic interactions predict protein‐small molecule and protein‐DNA/RNA affinities with similar accuracy

“…Affinity is commonly measured using the dissociation constant [Kd or pKd = −log(Kd)], which is the ligand concentration at which half the protein in solution is bound to ligand at equilibrium . Predicting molecular binding affinity from structural complexes has been investigated for decades, due to its fundamental importance in biochemistry and applications to structure‐based drug development . Most approaches attempt to produce a quantitative mapping to binding affinity from features that can be derived from a protein−ligand structure .…”

“…Mapping atomic interactions to binding affinity is not trivial, and a variety of methods have been developed. These methods can be broadly classified into approaches that attempt to directly model the physical forces contributing to molecular binding and those that rely on statistical associations between combinations of atom−atom interactions and ligand affinity . Robust physics‐based methods such as molecular dynamics are able to infer information about changes in system energy and other factors and can produce highly accurate affinity prediction, albeit at the cost of increased computation time .…”

“…9-13 Predicting molecular binding affinity from structural complexes has been investigated for decades, due to its fundamental importance in biochemistry and applications to structure-based drug development. [14][15][16][17][18] Most approaches attempt to produce a quantitative mapping to binding affinity from features that can be derived from a protein2ligand structure. 9,18,19 Mapping atomic interactions to binding affinity is not trivial, and a variety of methods have been developed.…”

“…These methods can be broadly classified into approaches that attempt to directly model the physical forces contributing to molecular binding and those that rely on statistical associations between combinations of atom2atom interactions and ligand affinity. 14,18,[20][21][22][23][24][25] Robust physics-based methods such as molecular dynamics are able to infer information about changes in system energy and other factors and can produce highly accurate affinity prediction, albeit at the cost of increased computation time. 26 Statistical prediction methods are typically much faster than physics-based approaches and can generally predict protein-small molecule affinity with accuracy suitable for high-throughput drug screening and similar applications.…”