DIA_QC_report

The dia_qc script contains several sub-commands to generate QC and batch reports from DIA proteomics data. Reports from one or more Skyline documents are combined into a single sqlite database which is modified or read by subsequent commands.

To see all the available sub-commands run: dia_qc -h

usage: dia_qc <command> [<args>]

Available commands:
   parse                Generate QC and batch correction database from Skyline reports.
   report_convert       Convert Skyline report(s) to parquet format.
   metadata_convert     Convert metadata annotations to specified file format.
   validate             Validate Nextflow pipeline parameters.
   filter               Filter replicates in database.
   impute               Impute missing precursor and/or protein values.
   normalize            Perform DirectLFQ or median normalization on batch database.
   qc_qmd               Generate QC qmd report.
   batch_rmd            Generate batch correction rmd report.
   export_gene_matrix   Export PDC gene tables from batch database.
   db_export            Export selected table(s) from precursor database.

Tools to generate QC and batch reports from DIA proteomics data

positional arguments:
  command        Subcommand to run.

options:
  -h, --help     Show this help message and exit.
  -v, --version  Show program version and exit.

Installation

Clone the repository

git clone --recurse-submodules https://github.com/ajmaurais/DIA_QC_report

The python scripts in DIA_QC_report are bundled into a package which can be installed with pip.

cd DIA_QC_report
pip install .

In order to render the batch Rmarkdown report, you also need to install the R package in the rDIAUtils sub-repository.

cd DIA_QC_report/rDIAUtils
Rscript -e 'install.packages(".", type="source", repo=NULL)'

Parsing Skyline reports

The parse sub-command is used to combine .tsv reports exported from one or more Skyline documents into a sqlite database which is used in all subsequent steps. parse requires a precursor level and replicate level report from each Skyline document. The templates for both reports are in the resources subdirectory.

Batch and QC database schema

To create a new database:

dia_qc parse --projectName <document_1> <document_1_replicate_quality.tsv> <document_1_precursor_quality.tsv>

To add reports from additional documents to an existing database add the --overwriteMode=append flag:

dia_qc parse --projectName <document_2> --overwriteMode=append <document_2_replicate_quality.tsv> <document_2_precursor_quality.tsv>

Normalzation

The normalize sub-command performs median normalization at the precursor level and DirectLFQ normalization at the protein level. The normalized values are written directly to the database. The script modifies the database in place by modifying the normalizedArea column in the precursors table and the normalizedAbundance column in the proteinQuants table.

dia_qc normalize data.db3

QC report

Generate qmd file

dia_qc qc_qmd --addStdProtein 'sp|P00924|ENO1_YEAST' data.db3

Render qmd to html

quarto render qc_report.qmd --to html

Batch correction report

Generate rmd file

dia_qc batch_rmd data.db3

Render rmd to html

Rscript -e 'rmarkdown::render("bc_report.rmd")'

Other sub-commands

Validate nextflow pipeline parameters

The validate sub-command can be used to validate parameters from a pipeline.config file or from command line arguments.

Export database tables to tsv file(s).

The db_export sub-command is used to export precursor, protein, or metadate tables from the batch/qc database to a tsv file.

dia_qc db_export data.db3

Convert between metadata formats

The metadata_conver sub-command is used to interconvert between all the metadata file formats which are supported by dia_qc parse.

Convert tsv to json

dia_qc metadata_convert -o json <metadata_tsv>

Convert tsv to Skyline annotation csv

dia_qc metadata_convert -o skyline <metadata_tsv>

Convert Skyline annotation csv to tsv

dia_qc metadata_convert -o tsv <annotation_csv>