Integrative workflows for metagenomic analysis.

Efthymios Ladoukakis,Fragiskos N Kolisis,Aristotelis A Chatziioannou

doi:10.3389/fcell.2014.00070

Abstract

The rapid evolution of all sequencing technologies, described by the term Next Generation Sequencing (NGS), have revolutionized metagenomic analysis. They constitute a combination of high-throughput analytical protocols, coupled to delicate measuring techniques, in order to potentially discover, properly assemble and map allelic sequences to the correct genomes, achieving particularly high yields for only a fraction of the cost of traditional processes (i.e., Sanger). From a bioinformatic perspective, this boils down to many GB of data being generated from each single sequencing experiment, rendering the management or even the storage, critical bottlenecks with respect to the overall analytical endeavor. The enormous complexity is even more aggravated by the versatility of the processing steps available, represented by the numerous bioinformatic tools that are essential, for each analytical task, in order to fully unveil the genetic content of a metagenomic dataset. These disparate tasks range from simple, nonetheless non-trivial, quality control of raw data to exceptionally complex protein annotation procedures, requesting a high level of expertise for their proper application or the neat implementation of the whole workflow. Furthermore, a bioinformatic analysis of such scale, requires grand computational resources, imposing as the sole realistic solution, the utilization of cloud computing infrastructures. In this review article we discuss different, integrative, bioinformatic solutions available, which address the aforementioned issues, by performing a critical assessment of the available automated pipelines for data management, quality control, and annotation of metagenomic data, embracing various, major sequencing technologies and applications.

Highlights

Metagenomics refers to the exhaustive study of a collection of genetic material, encompassing various genomes from a mixed community of organisms as defined from the National Human Genome Research Institute (Talking Glossary of Genetic Terms1)
Many bioinformatic pipelines have emerged that aim to address these issues through the provision of automated workflows and user friendly interfaces, in an effort to simplify the analytical procedure as much as possible, and minimize the entry barrier concerning the familiarization of the user with advanced programming or computational techniques
Each of these integrative analysis pipelines encapsulates a plethora of bioinformatic algorithms, seamlessly embedded into a multi-tasking framework that can address all aspects of a complete metagenomic analysis in an automated fashion

Summary

INTRODUCTION

A Galaxy workflow for metagenomic datasets was published (Kosakovsky Pond et al, 2009) that requires as input a single dataset of raw sequencing reads and performs an automated series of analyses exploiting specific integrated tools Those analyses include: (i) quality control and filtering of the reads (custom tool), (ii) text editing and data format converting (custom tools), (iii) homology search against NCBI-nt database (Megablast, Altschul et al, 1990), (iv) taxonomic analysis (custom tools), and (v) visualization of results (custom tools). The comparative data analysis suite contains (i) profile-based selection tools, (ii) gene neighborhood analysis tools, and (iii) multiple sequence alignment tools that can elucidate the gene content and phylogenetic profile of any metagenomic sample This platform constitutes a very robust and user friendly system for publishing and managing a user’s (meta) genome via its web server’s graphical user interface (GUI) as well as performing further functional annotation on it, while exploiting their cloud infrastructure. The most advanced user will find that it is a great solution for the conduct of complete and fully automated metagenomic analyses on a local server with dedicated resources

Findings

DISCUSSION

CONCLUSIONS