Chris Tomkins-Tinch
Chris Tomkins-Tinch
Add `dustmasker` as a tool to viral-ngs for pre-processing of metagenomics databases (removal of low-complexity sequences). It's part of the blast toolset so the `Tool` subclass should perhaps be added...
In a few places we use `.sam` intermediary files where we could use `.bam` files. The latter take a bit more IO/CPU time with the advantage of better compression ratio....
kjsiddle reports an issue with the feature table annotation transfer in the WDL-based GenBank preparation workflow: "tbl files for multiple segment genomes are taken in numerical order which can cause...
In the calls to build the viral-ngs conda package (to test, and to deploy), the `conda build` [commands](https://github.com/broadinstitute/viral-ngs/blob/master/travis/build-conda.sh#L48) should be wrapped with a retry loop, as the conda build operation...
Spades produces many short contigs. It may be helpful to have an optional parameter for `metagenomics.py::assemble_spades`, perhaps `--minContigLength`, to only write contigs greater in length than a specified value.
pipeline currently hangs as the pipe for the second fastq never emits output
The context manager, `util.file.open_or_gzopen()`, should perhaps be extended to support bzip2-compressed files. This can be done by incorporating the module `bz2`, which is part of the Python standard library for...
Currently Snakemake removes remote files when no remaining rules require the files. There should be an option (boolean in the `config.yaml` file?) to optionally preserve files downloaded by the pipeline...
The scaffolding stage does not always succeed for diverse genomes such as HIV or influenza. The parameter space of the involved tools (currently `nucmer`, part of [`MUMmer`](http://mummer.sourceforge.net/)) should be explored...
From @yesimon: > One thing that could save a lot of time is removing the dynamic tool install on `__init__` for things instances like `tools.Picard.SomeTool()`. Each time you create the...