title	PRIDE Submission Workflow
date	today
date-format	YYYY-MM-DD
format	typst output-file README.pdf

PRIDE upload

Instructions and templates for preparing a submission to PRIDE / ProteomeXchange.

See https://www.ebi.ac.uk/pride/markdownpage/submitdatapage for the PRIDE submission website, and Submission_Tutorial.pdf for the downloaded tutorial (v2.3.4) this template is based on.

Workflow

1. Fill out PRIDE_submission_template.xlsx with the project's metadata (see sheets below). Copy the template to a new file per project (e.g. PRIDE_submission_<project>.xlsx) rather than editing the template in place.

2. List the files and compute checksums with the scripts in tools/ (see Repository contents): export the file lists, curate the Files sheet, then run the checksums. Checksumming the raw files can take up to a day, so consider starting it well before you intend to submit. Cached results mean an interrupted or repeated run only processes new or changed files.

3. Generate the SDRF file in the Excel template:

   1. Fill in one row per raw file in the SDRF_Metadata sheet (input sheet with ontology hints and guideline links). The SDRF_tsv sheet is the export view of these values. Column order matters: the characteristics[...] columns (including characteristics[material type]) come first, then assay name and technology type, then the comment[...] columns, with comment[data file] and the factor value[...] columns last. The template combines several SDRF templates (base, sample-metadata, human, ms-proteomics; optionally clinical-metadata and dia-acquisition), listed with links in the Instructions sheet. Trim the columns to those you use, and see the notes in rows 1 and 3 of SDRF_Metadata for how to fill each one.

   2. Export the SDRF_tsv sheet (not SDRF_Metadata) with File > Save As > Text (Tab delimited) (*.txt) (active sheet only), then rename the extension to .sdrf.tsv.

   3. Check that the comment[data file] values match the file names in the Files sheet and in tools/checksum.txt exactly (including extension and case). PRIDE cross-references the SDRF against the submitted files, so a mismatch will cause problems.

   4. Validate the file with sdrf-pipelines, the official SDRF validator:

      pip install sdrf-pipelines
      parse_sdrf validate-sdrf --sdrf_file your_file.sdrf.tsv --template human --template ms-proteomics
      

      Pass one --template (-t) per template used in the comment[sdrf template] columns of the SDRF: the file is then validated against the union of all of them. Without --template only the default (ms-proteomics) is checked, so the human-specific columns would not be validated. For DIA data add --template dia-acquisition. parse_sdrf list-templates shows the available templates. Install sdrf-pipelines[ontology] to also validate ontology terms (or add --skip-ontology to skip that check). Fix any reported errors in the Excel sheet and re-export.

   Use lesSDRF (CompOmics) only as a lookup tool for the correct formatting of column values; do not build the file there. See lesSDRF_manual.pdf for the app's documentation, and the SDRF specification for the online reference on column names and allowed values.

4. Submit via the PX Submission Tool (px-submission-tool-2.11.5/): use the Dataset_Details, Files, File_Relationships, and Result_Metadata sheets to populate the tool's fields, upload the listed files, and attach the generated SDRF file, to create the PRIDE/ProteomeXchange submission.

Template structure (PRIDE_submission_template.xlsx)

Sheet	Purpose
Instructions	Overview of the workbook and notes on controlled-vocabulary (CV) fields (species/NEWT, tissue/BTO, instrument/PSI-MS, modifications/PSI-MOD or UNIMOD, cell type/CL, disease/DOID).
Dataset_Details	One set of project-level metadata: submission type (Complete/Partial), result file format, title, keywords, description, sample/data processing protocols, MS experiment type, submitter and lab head contact details.
Files	One row per file to submit, with its File type (RAW, PEAK, SEARCH, QUANT, FASTA, SPECTRUM_LIBRARY, GEL, OTHER, or RESULT for Complete submissions) and location notes.
File_Relationships	Many-to-one join table linking each RESULT (or, for Partial submissions, SEARCH) file to the RAW/PEAK/FASTA/etc. files it depends on. Every RAW, PEAK, SEARCH, and FASTA file in Files should appear here at least once.
Result_Metadata	Per-RESULT-file annotation: species, tissue, instrument, experimental factor, modifications/PTMs, cell type and disease (optional), quantification method.
SDRF_Metadata	Input sheet: one row per raw file/sample, with SDRF columns (source name, characteristics[...], assay name, comment[data file], etc.), plus ontology hints and links to the SDRF guidelines.
SDRF_tsv	Export view of SDRF_Metadata. Save this sheet as tab-delimited text and rename to .sdrf.tsv to produce the SDRF file.
Lists	Dropdown value lists used for data validation elsewhere in the workbook (submission type, result file format, file type).

Partial vs. Complete submissions: Partial submissions don't provide RESULT files in mzTab/mzIdentML form; use the SEARCH file (raw search engine output) as the "Result file" in File_Relationships instead, and leave Result file format blank in Dataset_Details.

Partial submission with MaxQuant — file type mapping:

File	File type (dropdown)
proteinGroups.txt, peptides.txt, evidence.txt (and other MaxQuant output tables)	RESULT
mqpar.xml (MaxQuant parameters / experiment design)	OTHER

(For Complete submissions the RESULT file must be mzTab or mzIdentML; Result file format is then set accordingly.)

Partial submission with DIA-NN — file type mapping:

File	File type (dropdown)	File list
*.mzML.quant (satisfies the Partial-submission SEARCH requirement)	SEARCH	SEARCH_files.txt
report.tsv / report.parquet (main report), report.pg_matrix.tsv, report.pr_matrix.tsv, report.gg_matrix.tsv, report.unique_genes_matrix.tsv	QUANT	DIANN_QUANT_files.txt
report.stats.tsv, report.log.txt	OTHER	DIANN_other_files.txt
*.speclib, report-lib.parquet (spectral library)	SPECTRUM_LIBRARY	SPECTRUM_LIBRARY_files.txt
FASTA database	FASTA	FASTA_files.txt

Note: this mapping is provisional. It follows PRIDE's file type definitions (SEARCH = search engine/pipeline output, required for Partial submissions; QUANT = quantification output) but PRIDE does not list DIA-NN explicitly. The .quant files are used as the SEARCH files; the main DIA-NN report is submitted as QUANT. Revise it if PRIDE support or a reviewer indicates different types, and keep the Files sheet and tools/Export-FileLists.ps1 in sync.

Repository contents

• PRIDE_submission_template.xlsx — blank template, copy this per project.

• tools/ — drop-in folder for the machine where the data lives. Copy the whole folder into the project's data directory and run the launchers by double-clicking; both scripts operate on the parent directory of the tools folder:

  • Export-FileLists.bat / Export-FileLists.ps1 — step 1: scans recursively for proteomics files of interest (.raw, .mzML, .quant search files, DIA-NN output (DIANN_QUANT_files.txt, DIANN_other_files.txt) and spectral libraries, sdrf.tsv/sdrf.txt or *.sdrf.tsv/*.sdrf.txt if already exported, MaxQuant output tables, mqpar.xml, .fasta, …) and writes one .txt per category into tools/_filelists/, grouped by folder (folder path, blank line, file names; two blank lines between folders). Copy names into the template's Files sheet and manually curate which files are actually submitted.
  • run_checksums.bat / make_checksums.py — step 2, after curating the Files sheet: paste the curated file names into tools/checksum_input.txt (one per line; folder-path lines copied from the file lists tell it where the names below are) and run.
    • Finds each file, computes SHA-1 checksums in parallel (Python 3.6+ standard library only), and writes tools/checksum.txt (file name<TAB>sha1, the format PRIDE wants).
    • Never guesses: if a name is missing, matches more than one file, or is listed twice, it lists the problem lines and writes nothing.
    • Prints a progress line as each file finishes.
    • Saves results to tools/checksum_cache.tsv, so files unchanged since an earlier run (same path, size and modification time) are not checksummed again, and an interrupted run resumes.
    • Keep make_checksums.py compatible with Python 3.6: don't let a code formatter rewrite it for newer Python. (This is because 3.6 is what is available and I don't have admin rights...)

  Both scripts are self-contained: they only read the parent directory, and all output stays inside tools/.

• px-submission-tool-2.11.5/ — the PX Submission Tool used to build and upload the submission.

• README.pdf — rendered copy of this README. To regenerate it, open README.md in Positron and run Quarto: Render Document from the command palette, or run quarto render README.md in a terminal.

• Submission_Tutorial.pdf, lesSDRF_manual.pdf — reference documentation (lookup only; the SDRF is built in the Excel template). Online SDRF reference: https://sdrf.quantms.org/index.html

Example tools/checksum_input.txt

Blocks can be pasted straight from the _filelists/*.txt files. A folder-path line applies to the file names below it; a name without a folder line (like uniprot.fasta here) is looked up anywhere in the project folder and must match exactly one file. Blank lines and # lines are ignored.

# Raw files
D:\Projects\Colonoids\Raw Data\Batch1

sample1.raw
sample2.raw


D:\Projects\Colonoids\Raw Data\Batch2

sample3.raw

# MaxQuant results (folder line needed: several runs
# have a proteinGroups.txt)
D:\Projects\Colonoids\MQ\run2\combined\txt

proteinGroups.txt
peptides.txt
evidence.txt

uniprot.fasta

Resulting tools/checksum.txt (file name, tab, SHA-1; in the same order as the input):

sample1.raw	764c16af46dd4f15edb05ecc5595b50cbe3714ea
sample2.raw	854d675a26dc0254a07b5725ba71242555863eb5
sample3.raw	4052402ba9118e4b5222aad52b7c87e949382b22
...