Releases: bonsai-team/matam
Releases · bonsai-team/matam
v1.1.0
Major changes
- Replace Sumaclust by VSearch
v1.0.0
Major changes
- Replace ComponentSearch (java) by ComponentSearch2 (c++)
Minor changes
- Externalise conda recipe to the following repository: https://github.com/bonsai-team/conda_recipes
- Add a new option to MATAM allowing the user to run only the first step of the pipeline
v0.9.9
v0.9.8
Major changes
- Contigs binning during scaffolding is now deactivated by default. This is still an experimental option.
- Updated Docker url
- Final assembly file is now explicit: "final_assembly.fa"
Minor changes
- Python sub-processes are now unbuffered
- Cleaned ovgraphbuild output
- Various cleaning and improvement to matam_assembly.py
v0.9.7
v0.9.6
v0.9.5
v0.9.4
Major changes
- Updated default DB to SILVA 128
- Default DB is now downloaded from bioinfo.lifl.fr and does not use git LFS anymore
- Added a new step to filter out references with a mapping coverage too high (most prevalent species)
- Improved Ns management during database preprocessing
Minor changes
- Corrected many bugs
- Improved output infos
- Bamtools and SGA are now compiled with the gcc version from the current user bash
- Performances improvement and better resources management
- Additional non-essential scripts
v0.9.3
Major Changes
- Cleaned and simplified ovgraphbuild code. No functional changes.
- Changed componentsearch default filtering parameters in matam_assembly.py
- matam_assembly.py now output assemblies stats at the end of the run
- Contigs assembly read correction is now an option (SGA was doing it by default) and is deactivated by default. This greatly improve assembly of low coverage species.
- Contig binning during scaffolding is now an option activated by default.
- Added options to matam_db_preprocessing.py to filter out too small or too long ref sequences
Minor Changes
- Added an option to sga_assembly.py to deactivate read correction and filtering
- Now doesnt crash if one of the assembly file is empty
- Added new script to compute fasta sequences length
- Added new script restore the .1 and .2 names to a paired fastq file missing them
v0.9.2-beta
Major changes
- Introduced compatible contigs binning in scaffolding step. This should prevent creating consensus scaffolds when contigs from 2 individuals/species map on the same reference.