Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
59 commits
Select commit Hold shift + click to select a range
0aa06ac
docs: clarify DIA-NN container licensing + local build; add 2.5.0 row
ypriverol May 30, 2026
0ce090d
chore: bump pipeline version to 2.2.0dev
ypriverol May 30, 2026
7fb4ca1
create_input_channel: allow empty FixedModifications
ypriverol May 31, 2026
23fe75e
fix(test_dia_dotd): correct PXD065380 database FASTA path
ypriverol Jun 2, 2026
42a4f27
fix(test_dia_dotd): correct PXD065380 database FASTA path
ypriverol Jun 2, 2026
e2c31a7
feat(qpx): flatten QPX export output into qpx/ (drop qpx_output/ subf…
ypriverol Jun 2, 2026
07006e7
chore(qpx): bump pin to bigbio/nf-modules main (qpx_output/* output)
ypriverol Jun 2, 2026
bc9569f
chore: restore trailing newline in modules.json (pre-commit end-of-fi…
ypriverol Jun 2, 2026
9fec585
docs(spec): version-dependent precursor q-value default (DIA-NN <2.5=…
ypriverol Jun 2, 2026
6e1f0c4
docs(spec): broaden to DIA-NN recommended defaults (version q-value +…
ypriverol Jun 2, 2026
fcbffd4
docs(spec): make user-provided precursor q-value override a hard, tes…
ypriverol Jun 2, 2026
912c1c8
Merge pull request #97 from bigbio/docs/diann-container-licensing
ypriverol Jun 2, 2026
1393111
feat(diann): version-aware precursor q-value + disable fast-mode cali…
ypriverol Jun 2, 2026
9a18326
Merge remote-tracking branch 'origin/dev' into feat/version-precursor…
ypriverol Jun 2, 2026
2a80dfe
Merge remote-tracking branch 'origin/dev' into fix/qpx-flatten-output
ypriverol Jun 2, 2026
d8419d5
docs(changelog): note QPX output flattening under 2.2.0dev
ypriverol Jun 2, 2026
245c2ba
Merge pull request #98 from bigbio/feat/allow-empty-fixed-modifications
ypriverol Jun 2, 2026
1578f5d
Merge pull request #100 from bigbio/fix/qpx-flatten-output
ypriverol Jun 2, 2026
e08eb68
Merge remote-tracking branch 'origin/dev' into feat/version-precursor…
ypriverol Jun 2, 2026
b49e9ce
Merge pull request #101 from bigbio/feat/version-precursor-qvalue
ypriverol Jun 2, 2026
e178a50
docs(spec): DIA-NN Enterprise 2.5.1 support (--kb, --license, enterpr…
ypriverol Jun 3, 2026
9bd9950
feat(diann): DIA-NN Enterprise 2.5.1 support (--kb knowledge base + -…
ypriverol Jun 3, 2026
ea60196
fix(diann-enterprise): align container tag to ghcr.io/bigbio/diann-en…
ypriverol Jun 3, 2026
6695c86
feat(diann-enterprise): enable Knowledge Base (--kb) by default under…
ypriverol Jun 3, 2026
2d61859
Merge pull request #102 from bigbio/feat/diann-enterprise
ypriverol Jun 3, 2026
7b17692
docs: clarify params.json vs SDRF-derived parameters
ypriverol Jun 3, 2026
8d72766
feat(diann): add academic DIA-NN 2.5.1 version profile (diann_v2_5_1)
ypriverol Jun 3, 2026
036b82d
Merge pull request #103 from bigbio/docs/clarify-params-vs-sdrf
ypriverol Jun 3, 2026
bc1e458
fix(diann): wire --kb (Enterprise Knowledge Base) into the search passes
ypriverol Jun 8, 2026
fcb0a77
fix(diann): use QuantUMS by default for DIA-NN >= 1.9.2
ypriverol Jun 8, 2026
af2847c
fix(diann): default --quant-sel-runs 20 and gate QuantUMS tuning flag…
ypriverol Jun 8, 2026
61d1d01
fix(diann): read optimised mass accuracy from PRELIMINARY logs for DI…
ypriverol Jun 8, 2026
b733069
fix(diann): mass-acc defaults 20/10 and SDRF>defaults when not auto-c…
ypriverol Jun 8, 2026
76c25cb
fix(diann): use --no-prot-inf in FINAL_QUANTIFICATION
ypriverol Jun 8, 2026
a506afa
fix(diann): fall back to SDRF/params when mass-acc calibration can't …
ypriverol Jun 8, 2026
4596bc3
docs(diann): record decision to keep pg_level default at 2
ypriverol Jun 8, 2026
4057d5e
fix(diann): register --kb + QuantUMS flags in BlockedFlags, sync sche…
ypriverol Jun 8, 2026
0d65a6d
Merge pull request #104 from bigbio/feat/bug-fixes-from-benchmarks
ypriverol Jun 9, 2026
7828a87
chore: bump sdrf-pipelines 0.1.4 -> 0.1.5 in SDRF_PARSING
ypriverol Jun 8, 2026
e7e6faa
Merge pull request #105 from bigbio/chore/bump-sdrf-pipelines-0.1.5
ypriverol Jun 9, 2026
71e1836
fix(diann): use toList() not collect() for per-run mass-acc triples
ypriverol Jun 9, 2026
71d1bac
Merge pull request #106 from bigbio/fix/dia-mass-acc-tolist
ypriverol Jun 9, 2026
569d205
feat(diann): make FINAL protein inference configurable (relaxed_prote…
ypriverol Jun 10, 2026
a2b8d48
docs: drop stale diann_extra_args references (use extra_args)
ypriverol Jun 10, 2026
470282d
feat(diann): add --strip_unknown_mods to recover hydroxyproline / non…
ypriverol Jun 10, 2026
528f67a
feat(diann): protein inference defaults to DIA-NN standard; relaxed/n…
ypriverol Jun 10, 2026
c9f974d
Merge pull request #109 from bigbio/fix/drop-diann-extra-args
ypriverol Jun 10, 2026
a83b70e
Merge pull request #108 from bigbio/feat/strip-unknown-mods
ypriverol Jun 10, 2026
f5205c3
Merge remote-tracking branch 'origin/dev' into feat/configurable-fina…
ypriverol Jun 10, 2026
3a42660
fix: drop re-introduced diann_extra_args (removed in favor of extra_a…
ypriverol Jun 10, 2026
e56a802
style(diann): drop verbose protein-inference comment in FINAL_QUANTIF…
ypriverol Jun 10, 2026
7fe7096
style(diann): drop inline comments that restate param descriptions
ypriverol Jun 10, 2026
89200b4
docs: document relaxed_prot_inf / no_prot_inf (usage table + CHANGELOG)
ypriverol Jun 10, 2026
47f349f
docs+blocking: clarify protein-inference params + tidy BlockedFlags
ypriverol Jun 10, 2026
b31a14b
style: trim verbose protein-inference comments in BlockedFlags to one…
ypriverol Jun 10, 2026
67f4bb3
docs: expand Protein inference section with DIA-NN's own examples
ypriverol Jun 10, 2026
39b0cb0
style: prettier-format docs/usage.md (CI pre-commit)
ypriverol Jun 10, 2026
de98e98
Merge pull request #107 from bigbio/feat/configurable-final-prot-inf
ypriverol Jun 10, 2026
f4a1f41
release: prepare 2.2.0 (Chongqing)
ypriverol Jun 16, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .nf-core.yml
Original file line number Diff line number Diff line change
Expand Up @@ -28,4 +28,4 @@ template:
name: quantmsdiann
org: bigbio
outdir: .
version: 2.1.0
version: 2.2.0
4 changes: 2 additions & 2 deletions AGENTS.md
Original file line number Diff line number Diff line change
Expand Up @@ -449,6 +449,6 @@ nextflow clean -f

---

**Last Updated**: 2026-05-08
**Pipeline Version**: 2.1.0
**Last Updated**: 2026-06-16
**Pipeline Version**: 2.2.0
**Minimum Nextflow**: 25.10.4
31 changes: 31 additions & 0 deletions CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,37 @@
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

## [2.2.0] bigbio/quantmsdiann — Chongqing - 2026-06-16

### `Added`

- New `--strip_unknown_mods` parameter (default `false`) that adds `--strip-unknown-mods` to the DIA-NN steps. Some declared variable modifications are not built into the DIA-NN deep-learning predictor — for example **Oxidation on proline**. During in-silico library generation DIA-NN **silently skips** those precursors (logged as `skipping N precursors, unrecognised modifications`), so they never enter the library and are never identified. Enabling `--strip_unknown_mods` forces DIA-NN to predict spectra/RTs/IMs for them so they are retained and searchable; it is a no-op when every declared modification is already recognised. Documented under [Where each GUI flag goes](docs/usage.md) and [Common pitfalls](docs/usage.md).
- New protein-inference parameters `relaxed_prot_inf` and `no_prot_inf` (both default `false`, mutually exclusive). By default the pipeline now uses DIA-NN's **standard** protein inference (no flag) in INDIVIDUAL_ANALYSIS + FINAL_QUANTIFICATION, matching DIA-NN's documented default. `relaxed_prot_inf` adds `--relaxed-prot-inf` (FragPipe/Spectronaut-style grouping); `no_prot_inf` adds `--no-prot-inf` (reuse the empirical-library inference). This replaces the previously hardcoded `--relaxed-prot-inf` (INDIVIDUAL_ANALYSIS) and `--no-prot-inf` (FINAL_QUANTIFICATION).
- DIA-NN **2.5.1** (academic) version profile `-profile diann_v2_5_1` (container `ghcr.io/bigbio/diann:2.5.1`).
- **DIA-NN Enterprise (2.5.1) support** via `-profile diann_v2_5_1_enterprise` (container `ghcr.io/bigbio/diann-enterprise:2.5.1`). New `--enable_kb` flag adds the Enterprise Knowledge Base (`--kb`) to the first-pass search to boost identifications (mainly human data); it is gated to the Enterprise build and **on by default** under the `diann_v2_5_1_enterprise` profile (disable with `--enable_kb false`). New `--diann_license <file>` stages the Enterprise license key into each DIA-NN step as `--license`, with fallback to a key bundled next to the binary when unset. The license key is a per-user secret and is never committed or bundled into the shared image.

### `Changed`

- Removed the stale `diann_extra_args` references. The parameter was renamed to `extra_args`, but the documented worked example still used the old name — which is silently ignored, so flags such as `--smart-profiling`/`--peak-center` were dropped. The worked example now uses `extra_args`, and passing `diann_extra_args` is flagged by nf-schema parameter validation as unrecognised. Use `extra_args` (or, better, the dedicated parameters where they exist).
- Documentation: clarified DIA-NN container licensing. Only DIA-NN 1.8.1 is publicly redistributable (pulled from BioContainers as the default); releases from 1.9 onward must be built locally from the [quantms-containers](https://github.com/bigbio/quantms-containers) recipes and tagged `ghcr.io/bigbio/diann:<version>` so the matching `-profile diann_v<version>` resolves them. Added the missing DIA-NN 2.5.0 row to the README support table.
- QPX export now publishes the Parquet dataset files and the `.h5mu` MuData file directly under `results/qpx/`, removing the intermediate `qpx/qpx_output/` subfolder. The `bigbio/qpx` module emits the dataset as `qpx_output/*` files (bigbio/nf-modules#39) so `publishDir` can flatten them.
- `--precursor_qvalue` default is now **version-aware**: unset resolves by `--diann_version` to `0.01` (1%) for DIA-NN < 2.5 and `0.05` (5%) for >= 2.5, matching DIA-NN's recommended precursor q-value. An explicit `--precursor_qvalue` always overrides and is never replaced by the version default. Applies to both the DIA-NN main report (`--qvalue`) and the MSstats input; the matrix thresholds (`--matrix_qvalue`, `--matrix_spec_q`) are unchanged.
- `--performance_mode` now defaults to **`false`**. The DIA-NN calibration speed flags `--min-corr 2 --corr-diff 1 --time-corr-only` can drop identifications on some data (per DIA-NN guidance) and are now opt-in. `--quick_mass_acc` is unchanged (still `true`).
- **QuantUMS** is now used by default for DIA-NN >= 1.9.2 (`--quantums true`, DIA-NN's recommended quantification). Set `--quantums false` to fall back to legacy quantification (adds `--direct-quant`). New `--quantums_sel_runs` (default `20`, maps to `--quant-sel-runs N`) auto-selects N runs to optimise QuantUMS parameters, keeping QuantUMS fast on large cohorts; the QuantUMS tuning flags are gated by DIA-NN version and only applied when QuantUMS is active.

### `Fixed`

- **Multi-residue modification sites are no longer silently dropped** (via the sdrf-pipelines 0.1.5 bump below). `parse_sdrf convert-diann` in 0.1.4 emitted multi-residue sites comma-separated (e.g. `Phospho,79.966331,S,T,Y`), which DIA-NN truncated to the first residue only (`S`) — dropping pT/pY (confirmed on PXD049692) and any modification spanning multiple residues. 0.1.5 concatenates sites (`S,T,Y` → `STY`, `M,P` → `MP`).
- Mass accuracy handling for DIA-NN >= 2.5.0: when `--mass_acc_automatic` is set, the optimised MS1/MS2 mass accuracy is now read back from the `PRELIMINARY_ANALYSIS` logs and reused in later passes. When calibration can't be parsed, the pipeline falls back to the SDRF-declared tolerances and then to the `--mass_acc_ms2`/`--mass_acc_ms1` defaults (now `20`/`10` ppm). The fallback order is `calibration > SDRF ppm > defaults` (auto on) or `SDRF ppm > defaults` (auto off).
- Per-run mass-accuracy triples are now collected with `toList()` instead of `collect()`, fixing channel cardinality when staging the per-run values.
- The Enterprise Knowledge Base (`--kb`) and the QuantUMS tuning flags are now registered in the centralized `lib/BlockedFlags.groovy` registry so they can't be smuggled in twice via `--extra_args`, and the schema defaults were synced to match.

### `Dependencies`

| Dependency | Old version | New version |
| ---------------- | ----------- | ----------- |
| `sdrf-pipelines` | 0.1.4 | 0.1.5 |

## [2.1.0] bigbio/quantmsdiann — Sao Pablo - 2026-05-05

### `Added`
Expand Down
36 changes: 30 additions & 6 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -36,15 +36,39 @@ The pipeline takes [SDRF](https://github.com/bigbio/proteomics-metadata-standard

## Supported DIA-NN Versions

| Version | Profile | Container | Key features |
| --------------- | -------------- | ------------------------------------------ | ---------------------------------------------- |
| 1.8.1 (default) | `diann_v1_8_1` | `docker.io/biocontainers/diann:v1.8.1_cv1` | Core DIA analysis, TSV output |
| 2.1.0 | `diann_v2_1_0` | `ghcr.io/bigbio/diann:2.1.0` | Native .raw support, Parquet output |
| 2.2.0 | `diann_v2_2_0` | `ghcr.io/bigbio/diann:2.2.0` | Speed optimizations (up to 1.6x on HPC) |
| 2.3.2 | `diann_v2_3_2` | `ghcr.io/bigbio/diann:2.3.2` | DDA support (beta), InfinDIA, up to 9 var mods |
| Version | Profile | Container | Key features |
| ---------------- | ------------------------- | ------------------------------------------ | ----------------------------------------------------------------- |
| 1.8.1 (default) | `diann_v1_8_1` | `docker.io/biocontainers/diann:v1.8.1_cv1` | Core DIA analysis, TSV output |
| 2.1.0 | `diann_v2_1_0` | `ghcr.io/bigbio/diann:2.1.0` | Native .raw support, Parquet output |
| 2.2.0 | `diann_v2_2_0` | `ghcr.io/bigbio/diann:2.2.0` | Speed optimizations (up to 1.6x on HPC) |
| 2.3.2 | `diann_v2_3_2` | `ghcr.io/bigbio/diann:2.3.2` | DDA support (beta), InfinDIA, up to 9 var mods |
| 2.5.0 | `diann_v2_5_0` | `ghcr.io/bigbio/diann:2.5.0` | +70% protein IDs, DL model selection flags |
| 2.5.1 | `diann_v2_5_1` | `ghcr.io/bigbio/diann:2.5.1` | Academic build of DIA-NN 2.5.1 |
| 2.5.1 Enterprise | `diann_v2_5_1_enterprise` | `ghcr.io/bigbio/diann-enterprise:2.5.1` | Knowledge Base (`--enable_kb`), extra QC metrics. Needs a license |

Switch versions with e.g. `-profile diann_v2_2_0,docker`. See the [DIA-NN Version Selection](docs/usage.md#dia-nn-version-selection) guide and [full parameter reference](docs/parameters.md) for details.

> [!IMPORTANT]
> **DIA-NN licensing.** Only **DIA-NN 1.8.1** is redistributable and is pulled
> automatically from the public BioContainers image
> (`docker.io/biocontainers/diann:v1.8.1_cv1`), so the default profile works out
> of the box. The DIA-NN license does **not** permit redistribution of releases
> from 1.9 onward, so the `ghcr.io/bigbio/diann:*` images above are **not
> public** — with a valid DIA-NN download you build them locally from the
> [`quantms-containers`](https://github.com/bigbio/quantms-containers) recipes:
>
> ```bash
> git clone https://github.com/bigbio/quantms-containers
> cd quantms-containers/diann-2.5.0
> docker build -t ghcr.io/bigbio/diann:2.5.0 . # Docker
> # or, for Singularity/Apptainer:
> singularity build diann-2.5.0.sif docker-daemon://ghcr.io/bigbio/diann:2.5.0
> ```
>
> The image **must** be tagged `ghcr.io/bigbio/diann:<version>` (matching the
> `Container` column) for the matching `-profile diann_v<version>` to pick it up
> automatically.

## Quick start

> [!NOTE]
Expand Down
14 changes: 14 additions & 0 deletions conf/diann_versions/v2_5_1.config
Original file line number Diff line number Diff line change
@@ -0,0 +1,14 @@
/*
* DIA-NN 2.5.1 container override (private ghcr.io)
* Academic build of DIA-NN 2.5.1 (the Enterprise build is a separate image,
* see conf/diann_versions/v2_5_1_enterprise.config and -profile diann_v2_5_1_enterprise).
*/
params.diann_version = '2.5.1'

process {
withLabel: diann {
container = 'ghcr.io/bigbio/diann:2.5.1'
}
}

// Container engine is selected via -profile (docker/singularity), not here
26 changes: 26 additions & 0 deletions conf/diann_versions/v2_5_1_enterprise.config
Original file line number Diff line number Diff line change
@@ -0,0 +1,26 @@
/*
* DIA-NN 2.5.1 Enterprise container override (private ghcr.io)
*
* Enterprise is a separate image from the academic build
* (academic 2.5.1, if shipped, is `ghcr.io/bigbio/diann:2.5.1`).
* The image bundles the Enterprise binary + the Knowledge Base model
* (models/base232.nnkb) but NO license key — supply the key at runtime
* with `--diann_license <file>` (or bake it next to the binary in a strictly
* private local build). The license key must never be committed or pushed.
*
* Enterprise adds the `--kb` Knowledge Base option (enable with `--enable_kb`),
* which boosts identifications, primarily on human data.
*/
params.diann_version = '2.5.1'
params.diann_enterprise = true
// Knowledge Base is the headline Enterprise feature, so it's on by default for
// this profile. Disable per-run with `--enable_kb false`.
params.enable_kb = true

process {
withLabel: diann {
container = 'ghcr.io/bigbio/diann-enterprise:2.5.1'
}
}

// Container engine is selected via -profile (docker/singularity), not here
2 changes: 1 addition & 1 deletion conf/modules/dia.config
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
========================================================================================
DIA-NN module options — all DIA pipeline steps
========================================================================================
Passes params.extra_args (formerly params.diann_extra_args) to each DIA-NN step via ext.args.
Passes params.extra_args to each DIA-NN step via ext.args.
Blocked-flag validation is performed in each module's script block,
so it applies regardless of whether ext.args is overridden by the user.
----------------------------------------------------------------------------------------
Expand Down
7 changes: 6 additions & 1 deletion conf/modules/shared.config
Original file line number Diff line number Diff line change
Expand Up @@ -84,11 +84,16 @@ process {
}

// QPX Parquet + MuData export (experimental, 2.1.0)
// Flatten the layout: the QPX dataset `*.parquet` files (emitted by the
// module as `qpx_output/*`) and the `*.h5mu` MuData file are published
// directly under `qpx/`, with no intermediate `qpx_output/` subfolder.
// `tokenize('/').last()` strips the `qpx_output/` prefix from the dataset
// files so everything lands flat in `qpx/`.
withName: '.*:QPX_EXPORT' {
publishDir = [
path: { "${params.outdir}/qpx" },
mode: 'copy',
saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
saveAs: { filename -> filename.equals('versions.yml') ? null : filename.tokenize('/').last() }
]
}
}
2 changes: 1 addition & 1 deletion conf/tests/test_dia_dotd.config
Original file line number Diff line number Diff line change
Expand Up @@ -27,7 +27,7 @@ params {

// Input data
input = 'https://raw.githubusercontent.com/bigbio/quantms-test-datasets/quantms/testdata/dia_ci_dotd/PXD065380.sdrf.tsv'
database = 'https://raw.githubusercontent.com/bigbio/quantms-test-datasets/quantms/databases/PXD065380.fasta'
database = 'https://raw.githubusercontent.com/bigbio/quantms-test-datasets/quantms/databases/PXD065380/PXD065380.fasta'

min_pr_mz = 350
max_pr_mz = 950
Expand Down
Loading
Loading