adapter-checks

A pipeline to validate read trimming preformance. Assumes paired end. Dependencies are the latest stable version of nextflow and conda (alternatively, BYO versions of the tools found in resources/environment.yml and remove all conda configuration in nextflow.config).

1. cutadapt is used to trim reads, given a known adapter sequence (--cutadapt_adapter). Other arguments can be provided with --cutadapt_args.
2. Independently, bbmerge is used to automatically detect adapter sequences from untrimmed data. Untrimmed files are subsetted to a fixed number of reads, configurable with --subsample_n.
3. bbmerge consensus sequences are compared to the adapter provided to cutadapt.
4. Consensus sequences are merged with a premade adapter list (can be altered via --adapters). The default uses the list provided by bbmap.
5. bbduk is used with the merged adapter list to filter for adapters. Two different sets of arguments are used, -bbduk_standard_args and -bbduk_lenient_args.
6. Random adapter sequences are generated and the same parameters are used with the random adapters as a comparison.
7. Report is generated. Most frequent matches are compared across each parameter combination.

Running

WORKFLOW=path/to/adapter-checks/

nextflow run path/to/adapter-checks \
    --input_fastq input_dir \
    --cutadapt_adapter 'CTGTCTCTTATACACATCT' \
    --cutadapt_args '-e 0.3' \
    --bbduk_standard_args 'ktrim=r mink=11 hdist=2 k=21 tbo' \
    --bbduk_lenient_args 'ktrim=r mink=11 hdist=1 k=23 tbo' \
    --output results_dir \
    -resume

Output

A report will be generated with some custom figures.

TODO: It would be nice to have those as part of a multiQC plugin, perhaps.