Datasets
The published study contains 13 integration tasks: five single-cell RNA tasks, six chromatin-accessibility tasks and two simulations. The tasks span 85 batches; individual task data and preprocessing remain separate.
Benchmark
scIB
scIB is an atlas-level single-cell integration benchmark study covering RNA, chromatin-accessibility and simulated datasets. It compares batch removal with preservation of biological variation. The scib Python package and scib-pipeline implement its evaluation workflow.
Sources (2)
scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100.Overview
Metrics
The metric module separates batch-correction and biological-conservation measures.
Sources
theislab/scib official source · Pinned README: package purpose; Metrics; Integration ToolsAllowed inputs
AnnData and integration outputs; label-dependent metrics additionally require biological labels.
Sources
theislab/scib official source · Pinned README: package purpose; Metrics; Integration ToolsEvaluation procedure diagram
How it worksEvaluation procedure
Conceptual workflow of the scIB benchmark study. Individual tasks, preprocessing choices and supervision settings remain separate.
Sources (2)
scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100.Source reviewed · Automated source review, 2026-09-19. All specifications and missing details
Results
Each comparison retains its reviewed evaluation scope, dataset and metric. Results are shown without a pooled ranking.
scIB official RNA metrics export · pancreas: NMI_cluster/label
NMI_cluster/label (score) · Higher values are better.
scIB official RNA metrics export · pancreas · pancreas
Evidence origin: Author-reported evaluation.
scIB: 3afbffd3674726e5146797be21cf6bd7470a2c5f · data/metrics.csv row 2, column 2 (NMI_cluster/label) through data/metrics.csv row 410, column 2 (NMI_cluster/label)- Published metric-score export; intended higher direction verified from pinned official consuming code. No additional min-max normalization or overall ranking. Dataset metadata reports total cells; per-result scoring coverage is unreported.
- Missing source cells and quarantined conflicts are recorded in acquisition and audit tables. Per-result scoring denominators may be unreported.
Comparison details and limitations
Complete selected source table is retained across source-order panels. These point estimates do not establish statistical significance or a universal ranking.
- Source-specific evaluation. No equivalence to other releases, protocols or model families is inferred.
Automated source review: 2026-09-19. Numerical source review does not establish independent reproduction.
Dots show point estimates. Whiskers show only explicitly defined uncertainty (standard deviation, standard error or a labelled interval); their definitions remain in Table. Unresolved uncertainty is not plotted. Differences do not establish statistical significance.
Showing 12 of 69 matching rows.
Tested configuration
0.2970.4750.6520.8291.01
Reported score- scgen_full · hvg, unscaled0.968
- scgen_full · full_feature, unscaled0.949
- fastmnn_embed · hvg, unscaled0.924
- scanorama_embed · hvg, unscaled0.922
- scanvi_embed · hvg, unscaled0.922
- fastmnn_full · hvg, unscaled0.921
- mnn_full · hvg, scaled0.918
- scvi_embed · hvg, unscaled0.914
- conos_knn · full_feature, unscaled0.914
- mnn_full · hvg, unscaled0.914
- combat_full · hvg, scaled0.912
- seuratrpca_full · hvg, scaled0.91
Methods and evaluation design
Procedure, tasks and evaluated configurations
How it works
Evaluation methodology
The published scIB study compares integration methods on 13 atlas-level tasks spanning RNA, chromatin-accessibility and simulated data. Each task defines its batches and biological annotations. Four preprocessing choices combine scaled or unscaled values with full features or highly variable features for the RNA comparisons. Metrics measure batch correction and biological conservation separately. Label-assisted methods must retain that additional-input distinction. The scib package computes metrics and the separate scib-pipeline coordinates integrations; neither software package is itself a fixed dataset or a universal train/test split.
Sources (2)
scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100.Evaluation design
Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.
These source-backed links do not make different protocols or scores interchangeable.
Baseline coverage
Reference methods help show what a model adds beyond simple controls. We track a null control and a conventional method for each protocol.
0 of 2 active baseline roles have published Rewire measurements in this release. Measurements on a selected protocol do not establish coverage of an entire suite.
Baseline status by linked protocol
- scIB official RNA metrics export · pancreas · 0/2 roles measured
Protocol coverage CSV · Model evaluation matrix · Source table · Release and checksums
Coverage is derived from release 2026-09-29-06401fd5b220. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.
Run this benchmark
Choose a concrete protocol before running an evaluation. Its inputs, split and scoring rules determine which results can be compared.
Official run instructions
Run the historical scIB pipeline on its supplied test-data configuration, beginning with a Snakemake dry run.
Checked against the official instructions on 2026-09-17. These commands have not been executed by rewire. Running them does not automatically reproduce the published scores.
Before you start
- Conda or Mambaforge/Mamba and Git; the official setup creates separate Python and R environments.
- The R 4.0 example uses scib-pipeline-R4.0 and scib-R4.0. Data must contain normalized log counts in adata.X and raw counts in adata.layers["counts"].
1. Check out the reviewed repository
Repository checkout wrapper: the detached revision selects the exact official source inspected for this guide.
theislab/scib-pipeline / README.md · Pinned repository revision; README.mdgit clone https://github.com/theislab/scib-pipeline.git cd scib-pipeline git checkout --detach e97631a478e063883cee5db0dcebbf42d8268ab02. Create and activate the documented environments
The setup invocation is official; environment activation selects the Python environment named in the README. Conda shell initialization is a prerequisite.
theislab/scib-pipeline / README.md · README.md lines 33–68bash envs/create_conda_environments.sh -r 4.0 conda activate scib-pipeline-R4.03. Generate the included test data
The subshell supplies the data-directory working context required by the official data README. Existing output is not overwritten by default.
theislab/scib-pipeline / data/README.md · data/README.md lines 1–10(cd data && python generate_data.py)4. Inspect scheduled work
Dry run: shows the jobs without executing integrations or metrics. The .yaml suffix is verified in the pinned repository; the README installation table contains a conflicting .yml spelling.
theislab/scib-pipeline / README.md; theislab/scib-pipeline / configs/test_data-R4.0.yaml · README.md lines 91–99; configs/test_data-R4.0.yamlsnakemake --configfile configs/test_data-R4.0.yaml -n5. Run the supplied test-data benchmark
Official example permits up to ten cores. Inspect the configuration first: it includes several Python and R integration methods, so this is not a single-method smoke test.
theislab/scib-pipeline / README.md; theislab/scib-pipeline / configs/test_data-R4.0.yaml · README.md lines 99–104; configs/test_data-R4.0.yamlsnakemake --configfile configs/test_data-R4.0.yaml --cores 10
Expected outputs
- data/adata_norm.h5ad from the test-data generator.
- Integration and metric outputs produced by the Snakemake targets selected in the configuration.
Scope and limitations
- The pipeline is tailored to the historical scIB study and the maintainers redirect general-purpose use to newer projects; this guide does not claim current Open Problems equivalence.
- The test-data configuration is not the original 85-batch study dataset.
- Old R/Python environment resolution and method-specific hardware support were not executed or verified.
- The inspected instructions do not establish a minimum RAM/VRAM requirement, wall-clock runtime or monetary cost; none is inferred.
Strengths, limitations and unresolved questions
Strengths and limitations
Strengths supported by sources
- Separate batch-removal and biological-conservation scores reveal their trade-off.
Sources
theislab/scib official source · Pinned README: package purpose; Metrics; Integration Tools
Limitations and conditions
- Batch mixing and preservation of biological variation must be considered together. Annotation-assisted methods and unsupervised methods receive different input information, and integration of known batches does not establish transfer to unseen batches.
Sources
scib primary benchmark evidence · Methods and Results: integration inputs, label use and biological-conservation metrics
Profile review details
The study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed.
Stable record: discovery-benchmark-scibSpecifications
Inputs, training, access and other details
Explanatory profile: source reviewed · Automated source review, 2026-09-19. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.
| Property | Description and evidence |
|---|---|
| Datasets | The published study contains 13 integration tasks: five single-cell RNA tasks, six chromatin-accessibility tasks and two simulations. The tasks span 85 batches; individual task data and preprocessing remain separate.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Splits | Each task integrates its specified input batches. This is not a single held-out-cell classifier split. Dataset, preprocessing and use of biological labels must be recorded for each compared method.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Metrics | The metric module separates batch-correction and biological-conservation measures.Sourcestheislab/scib official source · Pinned README: package purpose; Metrics; Integration Tools |
| Baselines | Listed integrations include Harmony, MNN/FastMNN, scVI/scANVI, Scanorama, BBKNN and Seurat.Sourcestheislab/scib official source · Pinned README: package purpose; Metrics; Integration Tools |
| Leakage controls | scIB evaluates integration of the supplied batches together. Some methods use cell-type labels and others do not; the paper reports this distinction. This transductive integration setting is not a held-out-cell classifier test, so supervised train/test leakage terminology cannot be applied without specifying the method.Sourcesscib primary benchmark evidence · Methods and Results: integration inputs, label use and biological-conservation metrics |
| Uncertainty | Uncertainty and scoring coverage must be recorded for the individual task and metric. A study-wide task or cell count does not establish the denominator or uncertainty of every method result. · Not reported in inspected sourcesSources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Entity type | Benchmark study and collection of integration tasks; the scib evaluator package and scib-pipeline are supporting software.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Organisms | The study includes human and mouse atlas tasks, alongside simulated tasks. Organism and cross-species composition belong to the individual task.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Assays | Single-cell RNA expression and chromatin accessibility, plus simulated expression data. Keep assay and feature representation attached to the task.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
| Allowed inputs | AnnData and integration outputs; label-dependent metrics additionally require biological labels.Sourcestheislab/scib official source · Pinned README: package purpose; Metrics; Integration Tools |
| Adaptation | Integration methods operate on each task’s supplied batches with their reported preprocessing and supervision settings. The evaluation software scores those outputs; fitting belongs to the tested method.Sources (2)scib primary benchmark evidence; theislab/scib / README.md · Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. |
Applicable tests and references
Applicability is distinct from a completed evaluation.
- Unintegrated expression reference · proposed
- scVI · Proposed association
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Papers and result coverage
Last literature check: 2026-09-17. Primary-source discovery and table/protocol screening; source checked is not independently reproduced. Raw acquisitions not automatically numerical publication approval.
| Paper or primary resource | Version | Reference |
|---|---|---|
| Benchmarking atlas-level data integration in single-cell genomics | PMC8748196 | Read source DOI: 10.1038/s41592-021-01336-8 |
Historical gaps recorded on 2026-09-17
The catalogue now holds 821 result rows for this benchmark. A note below about pending extraction describes the state on 2026-09-17 and may since have been answered by a later batch. The result rows and their sources are the current record.
- Bio-conservation and batch-removal components depend on datasets and preprocessing. Figures do not justify new exact numerical scores; official source-data matrix extraction remains pending. Broad batch integration is not identical to scIB or Open Problems.
Search and extraction details
source found structured extraction pending
Searches
- scIB primary paper benchmark results
Evidence locations
- Table1 tasks and integration-evaluation methods; result figures
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
38 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Diagram caption Conceptual workflow of the scIB benchmark study. Individual tasks, preprocessing choices and supervision settings remain separate. Individual claims | scib primary benchmark evidence Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: PMC8748196 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram caption Conceptual workflow of the scIB benchmark study. Individual tasks, preprocessing choices and supervision settings remain separate. Individual claims | theislab/scib / README.md Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Diagram steps
| scib primary benchmark evidence Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: PMC8748196 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Diagram steps
| theislab/scib / README.md Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram title Evaluation procedure Individual claims | scib primary benchmark evidence Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: PMC8748196 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram title Evaluation procedure Individual claims | theislab/scib / README.md Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Datasets The published study contains 13 integration tasks: five single-cell RNA tasks, six chromatin-accessibility tasks and two simulations. The tasks span 85 batches; individual task data and preprocessing remain separate. Individual claims | scib primary benchmark evidence Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: PMC8748196 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Datasets The published study contains 13 integration tasks: five single-cell RNA tasks, six chromatin-accessibility tasks and two simulations. The tasks span 85 batches; individual task data and preprocessing remain separate. Individual claims | theislab/scib / README.md Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Splits Each task integrates its specified input batches. This is not a single held-out-cell classifier split. Dataset, preprocessing and use of biological labels must be recorded for each compared method. Individual claims | scib primary benchmark evidence Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: PMC8748196 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Splits Each task integrates its specified input batches. This is not a single held-out-cell classifier split. Dataset, preprocessing and use of biological labels must be recorded for each compared method. Individual claims | theislab/scib / README.md Primary paper Abstract; Results: Single-cell integration benchmarking (scIB), Figure 1 and Table 1; Methods: Datasets and preprocessing; Code availability. Pinned scib README lines 7-33, 62-100. Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468 | source checked automated source review · 2026-09-19 Audit detailsThe study/package distinction, task inventory and revised procedure descriptions were checked by a separate AI-assisted reviewer against the primary paper and pinned package README. Other profile claims retain their existing reviews. No human sign-off or model reproduction is claimed. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
View linked audit checks and correction history
Release 2026-09-29-06401fd5b220 · Record review: discovered
7 source records and release history
- theislab/scib official source · Original source · cd67913396b4c0430710b3d90f1d1841f5fa4468
- Benchmarking atlas-level data integration in single-cell genomics · Original source · PMC8748196
- theislab/scib / README.md · Original source · Git commit cd67913396b4c0430710b3d90f1d1841f5fa4468
- theislab/scib-pipeline / README.md · Original source · Git commit e97631a478e063883cee5db0dcebbf42d8268ab0
- theislab/scib-pipeline / data/README.md · Original source · Git commit e97631a478e063883cee5db0dcebbf42d8268ab0
- theislab/scib-pipeline / configs/test_data-R4.0.yaml · Original source · Git commit e97631a478e063883cee5db0dcebbf42d8268ab0
- scib primary benchmark evidence · Original source · PMC8748196
Technical metadata and extraction receipts
Stable ID: discovery-benchmark-scib
- areas
- single-cell
- entity level
- suite
- scope note
- Specialist molecular or omics evaluation; protocol details require review before numerical comparison.
- task
- Single-cell integration evaluation
- version
- Not reported
- benchmark research
- review date: 2026-09-17; status: source_found_structured_extraction_pending; primary sources: evidence-expansion-p2-evidence-discovery-final-scib-f65dd8b63336; inspected locators: Table1 tasks and integration-evaluation methods; result figures; searched queries: scIB primary paper benchmark results; gaps: Bio-conservation and batch-removal components depend on datasets and preprocessing. Figures do not justify new exact numerical scores; official source-data matrix extraction remains pending. Broad batch integration is not identical to scIB or Open Problems.; claim scope: Primary-source discovery and table/protocol screening; source checked is not independently reproduced. Raw acquisitions not automatically numerical publication approval.
- historical missing metadata
- dataset release: unextracted; metric implementation: unextracted; split manifest: unextracted; version: unextracted
- metadata review scope
- historical_missing_metadata preserves the original discovery state. Current descriptive evidence and missingness are recorded in profile.facts; numerical-result review is separate.
- entity classification
- review date: 2026-09-17; rationale: The cited profile describes a collection of evaluation tasks or protocols; retain it as the top-level benchmark suite. Its datasets and individual protocols remain separate records.; source ids: src-discovery-theislab-scib; source locator: Pinned README: package purpose; Metrics; Integration Tools; ambiguities: None recorded
- run guide
- record id: discovery-benchmark-scib; summary: Run the historical scIB pipeline on its supplied test-data configuration, beginning with a Snakemake dry run.; status: source_reviewed_not_executed; prerequisites: Conda or Mambaforge/Mamba and Git; the official setup creates separate Python and R environments.; The R 4.0 example uses scib-pipeline-R4.0 and scib-R4.0. Data must contain normalized log counts in adata.X and raw counts in adata.layers["counts"].; steps: title: Check out the reviewed repository; shell: git clone https://github.com/theislab/scib-pipeline.git cd scib-pipeline git checkout --detach e97631a478e063883cee5db0dcebbf42d8268ab0; explanation: Repository checkout wrapper: the detached revision selects the exact official source inspected for this guide.; source ids: run-doc-scib-pipeline-readme-md-e97631a4; source locator: Pinned repository revision; README.md; title: Create and activate the documented environments; shell: bash envs/create_conda_environments.sh -r 4.0 conda activate scib-pipeline-R4.0; explanation: The setup invocation is official; environment activation selects the Python environment named in the README. Conda shell initialization is a prerequisite.; source ids: run-doc-scib-pipeline-readme-md-e97631a4; source locator: README.md lines 33–68; title: Generate the included test data; shell: (cd data && python generate_data.py); explanation: The subshell supplies the data-directory working context required by the official data README. Existing output is not overwritten by default.; source ids: run-doc-scib-pipeline-data-readme-md-e97631a4; source locator: data/README.md lines 1–10; title: Inspect scheduled work; shell: snakemake --configfile configs/test_data-R4.0.yaml -n; explanation: Dry run: shows the jobs without executing integrations or metrics. The .yaml suffix is verified in the pinned repository; the README installation table contains a conflicting .yml spelling.; source ids: run-doc-scib-pipeline-readme-md-e97631a4; run-doc-scib-test-config-yaml-e97631a4; source locator: README.md lines 91–99; configs/test_data-R4.0.yaml; title: Run the supplied test-data benchmark; shell: snakemake --configfile configs/test_data-R4.0.yaml --cores 10; explanation: Official example permits up to ten cores. Inspect the configuration first: it includes several Python and R integration methods, so this is not a single-method smoke test.; source ids: run-doc-scib-pipeline-readme-md-e97631a4; run-doc-scib-test-config-yaml-e97631a4; source locator: README.md lines 99–104; configs/test_data-R4.0.yaml; outputs: data/adata_norm.h5ad from the test-data generator.; Integration and metric outputs produced by the Snakemake targets selected in the configuration.; limitations: The pipeline is tailored to the historical scIB study and the maintainers redirect general-purpose use to newer projects; this guide does not claim current Open Problems equivalence.; The test-data configuration is not the original 85-batch study dataset.; Old R/Python environment resolution and method-specific hardware support were not executed or verified.; The inspected instructions do not establish a minimum RAM/VRAM requirement, wall-clock runtime or monetary cost; none is inferred.; source ids: run-doc-scib-readme-md-cd679133; run-doc-scib-pipeline-readme-md-e97631a4; run-doc-scib-pipeline-data-readme-md-e97631a4; run-doc-scib-test-config-yaml-e97631a4; review: method: official_repository_review; date: 2026-09-17
- run documentation
- record id: discovery-benchmark-scib; source ids: run-doc-scib-readme-md-cd679133; run-doc-scib-pipeline-readme-md-e97631a4; run-doc-scib-pipeline-data-readme-md-e97631a4; run-doc-scib-test-config-yaml-e97631a4; status: source_reviewed_not_executed; summary: Run the historical scIB pipeline on its supplied test-data configuration, beginning with a Snakemake dry run. Commands were source-reviewed only. The pipeline is tailored to the historical scIB study and the maintainers redirect general-purpose use to newer projects; this guide does not claim current Open Problems equivalence.; source locator: Pinned repository revision; README.md; README.md lines 33–68; data/README.md lines 1–10; README.md lines 91–99; configs/test_data-R4.0.yaml; README.md lines 99–104; configs/test_data-R4.0.yaml
Related records
- evaluates task: Batch integration
- part of: scIB official RNA metrics export · pancreas
- applicable to: Unintegrated expression reference
- applicable to: scVI
- subject: scIB: evaluates task Batch integration