README.md

Small bioinfo containers

These containers are small (filesize) images for genomics/transcriptomics. Each submodule provides one app.

If you find this useful in your research, please cite the DOI:

Currently we have builds for:

Alignment

• bschiffthaler/bowtie
• bschiffthaler/bowtie2
• bschiffthaler/bwa
• bschiffthaler/diamond
• bschiffthaler/gmap
• bschiffthaler/minimap2
• bschiffthaler/ncbi-blast
• bschiffthaler/star

Annotation

• bschiffthaler/sorffinder

Assembly

• bschiffthaler/flye
• bschiffthaler/spades
• bschiffthaler/wtdbg2
• bschiffthaler/stringtie

Assembly Evaluation

• bschiffthaler/gffcompare
• bschiffthaler/transrate

Data Retrieval

• bschiffthaler/basemount

K-mer counting

• bschiffthaler/jellyfish

Linux Pipeliners

• bschiffthaler/awk
• bschiffthaler/bioawk
• bschiffthaler/curl
• bschiffthaler/sed

Multiple Alignment

• bschiffthaler/clustalw

Population genomics

• bschiffthaler/stacks

Pseudo-alignment/quantification

• bschiffthaler/kallisto
• bschiffthaler/salmon

Pseudogene detection

• bschiffthaler/pseudopipe

QC/Read preprocessing

• bschiffthaler/demultiplex
• bschiffthaler/cutadapt
• bschiffthaler/fastqc
• bschiffthaler/fastq_screen
• bschiffthaler/multiqc
• bschiffthaler/sortmerna
• bschiffthaler/trim-galore
• bschiffthaler/trimmomatic
• bschiffthaler/kraken2

Toolboxes

• bschiffthaler/bbtools
• bschiffthaler/bcftools
• bschiffthaler/bedtools
• bschiffthaler/genometools
• bschiffthaler/gffread
• bschiffthaler/samtools
• bschiffthaler/htslib

Motivation

Docker is an excellent tool for bioinformatics as we often see users with very limited exposure to Linux trying to use software that can be complex to install. With reproducible research in mind, it also offers an excellent way to distribute exact versions of tolopython3 tamsl as a supplement to the materials and methods. This is currently severely hampered by multi-gigabyte images that provide a complex stack of libraries in order to execute a single binary. Further, users may want to keep several versions of tools for compatibility, but these libraries can quickly grow to serveral hundred GBs even with docker's smart storage drivers.

This project aims to build images that are as small as possible to allow docker to be used more effectively on research compute systems (small or large) and to enable better reproducible research so containers can be supplementary to materials and methods.

Usage

These containers are typically intended to be run as exemplified:

docker run \
  --user=${UID}:${GID} \
  -v $(pwd):$(pwd) -w $(pwd) --rm \
  bschiffthaler/bwa ... # Further argument passed to bwa

Alternatively there are simple bash scripts in the /scripts directory that wrap the docker images and allow users to call them very similaly to a native CLI application. As an example:

<kogia_root>/scripts/samtools index GRCh38.fa

Some containers may run differently, see READMEs in the sub-directories.

Contributing

Pull requests are very welcome. Contibuted Dockerfiles need to be based on alpine (with rare exceptions) and should follow the two image build procedure (see the bwa image as an example).

Thanks to all who have contributed so far: @nicolasDelhomme @krbe9202