Minor updates

Only minor updates:

• Reformatting and additional long format table to compare in silico results
• Prevent an bug in normalization of tables containing "-" in Sample ids
• MultiQC can be executed from the output of DADA2
• Readme was partly re-written, now with references of embedded tools

Major update

New features:

Simplification of DB processing script:

• The script used for dereplication of reference sequences and containerization of related taxa has been simplified. It now works or more databases than only EzbioCloud.
• EzBioCloud was reprocessed with this new script, as well as Silva (available in the Docker image).

New Decipher classifier:

• CA Murali et al. (2018) "IDTAXA: a novel approach for accurate taxonomic classification of microbiome sequences." Microbiome, doi:10.1186/s40168-018-0521-5.
• Fast, requiring less memory than RDP
• Base on a pre-compiled taxonomy (available in the Docker image)

ITS:

• Implemented unite taxonomic DB for fungi (https://dx.doi.org/10.15156/BIO/786334)
• available in the Docker image
• ITS or 16S must be defined in for config ITS_or_16S: parameter to enable primer trimming settings adapted to the variable length of ITS (for the cutadapat/DADA2 processing only).

Normalization:

• a set of rule and script in reads post-processing enables normalization of the reads thanks to functions from several packages (Aldex2, vegan, metagenomeSeq, edgeR).

Updated tools:

• Several tools have been updated to the last version (see .yml in env folder).

Files renaming

• Several files were renamed

Potential future dev:

• adaptation of PANDAseq/vsearch pipeline for ITS
• comparison of DADA2 and Qiime implementation of RDP classifier to keep only one
• Develop visualization interface for QC (for instance with instance with Django)
• Develop visualization interface for plotting (for instance with shiny)

In silico validation

• Thanks to the taxonomy matching and assemblies downloading abilities of https://github.com/metagenlab/assembly_finder, we have implemented In-silico validation.
  Here amplicons are extracted by simulate_PCR (https://doi.org/10.1186/1471-2105-15-237) and assigned taxonomically by our pipeline.

• Docker building and validation are now split but both executed from a bash script.

• RDP taxonomic classifier wrapped in DADA2 was implemented. If providing comparable results with the one implemented in Qiime, it would allow to get rid of Qiime, making the Docker build lighter.

Singularity compatible Dockerfile

• Dockerfile definition was corrected to be compatible with Singularity

First Docker release

• First Docker released (metagenlab/amplicon_pipeline:v.0.9.6-beta)
• Can accept SRA as input
• Structure of the pipeline significantly reviewed for simplification
• Qiime1 RDP assign to simple (not multi-threaded), it was a true multithreading and was prone to silent failure

Numerous minor changes

Merge pull request #12 from metagenlab/dev

Dev

After ANCOM and Gneiss implementation

Current state when data have been transferred to Burton for TMA analysis. A lot of modification occurred since last release:

• config file has been completed and cleaned. vsearch vs DADA2 and databases are now parameters in config.
• Gneiss and ANCOM stat tools from Qiime2 have been implemented.
• Phyloseq samples, taxa and features filtration have been implemented.

New release

v0.9.3-alpha

minor changes heatmap function

Current state R&D committee 12.2019

Pipeline version at time of submission of the report at the december 2018 R&D committee.

alpha-release

Alpha-release