scTab: Scaling cross-tissue single-cell annotation models
10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30
What this source contributed
Records in this release that cite this source. A record can cite several sources.
Results
Values as the source printed them. Compare results only under compatible protocols, datasets and metrics.
Claims
Statements this source supports, each with where in the source it was checked.
- Baseline: Transfer cell-type annotations to a new dataset: Baseline scmap-cell, linear SVM and logistic regression (ucc-research-source-abdelaal-2019-paper: Abstract and Background (manual annotation), Results (SVM across the pancreatic datasets); ucc-research-source-sctab-paper: Results and Methods (scaled logistic regression against the CellTypist reference))
- CellTypist — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- CIForm — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- CIForm: description: Description Transformer-based cell-type annotation model for single-cell RNA-seq, benchmarked against scTab. (ucc-research-source-sctab-paper: Results (CIForm benchmark paragraph))
- MLP — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- MLP: description: Description Multi-layer perceptron cell-type classifier previously proposed for single-cell annotation, used as a nonlinear baseline. (ucc-research-source-sctab-paper: Results (nonlinear models benchmarked against the linear baseline))
- Optimized linear logistic regression — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- Optimized linear logistic regression: description: Description Logistic-regression cell-type classifier implemented to train on the full corpus with a cross-entropy loss, as the study's optimised linear baseline. (ucc-research-source-sctab-paper: Results (scaling logistic regression to the full training data))
- Relevance of scTab deep-ensemble uncertainty ROC-AUC: known-type correct vs. incorrect (Group 1 vs. Group 2) to "Transfer cell-type annotations to a new dataset": Assessed by scTab deep-ensemble uncertainty ROC-AUC: known-type correct vs. incorrect (Group 1 vs. Group 2) (ucc-research-source-sctab-paper: Methods, "Uncertainty quantification for scTab model")
- Relevance of scTab deep-ensemble uncertainty ROC-AUC: known-type vs. absent-type (Group 1 vs. Group 3) to "Transfer cell-type annotations to a new dataset": Assessed by scTab deep-ensemble uncertainty ROC-AUC: known-type vs. absent-type (Group 1 vs. Group 3) (ucc-research-source-sctab-paper: Methods, "Uncertainty quantification for scTab model")
- Relevance of scTab Table 1a: seeded model fits to "Transfer cell-type annotations to a new dataset": Assessed by scTab Table 1a: seeded model fits (ucc-research-source-sctab-paper: Supplementary Table 1a; Results and Methods; ucc-research-source-sctab-supp: Supplementary Table 1a; Results and Methods; ucc-research-source-sctab-code: Supplementary Table 1a; Results and Methods)
- scGPT fine-tuned — scTab Table 1: facets.method_types: Facets: method types ["foundation_model"] (scTab Methods: fine-tuned scGPT and frozen scGPT embeddings; catalog-model-scgpt configurations are foundation_model)
- scGPT frozen embeddings + logistic regression — scTab Table 1: facets.method_types: Facets: method types ["foundation_model"] (scTab Methods: fine-tuned scGPT and frozen scGPT embeddings; catalog-model-scgpt configurations are foundation_model)
- scTab — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- scTab: description: Description Scalable implementation of the TabNet architecture adapted to single-cell data: expression is size-factor normalised and log1p transformed, and observation-wise feature attention selects the genes used for cell-type classification. (ucc-research-source-sctab-paper: Results (scTab architecture); Discussion (observation-wise feature attention))
- Source membership: ucc-research-protocol-sctab-donor-bootstrap: Part of scTab published annotation evaluation (Supplementary Tables 1,2,6)
- Source membership: ucc-research-protocol-sctab-resources: Part of scTab published annotation evaluation (Supplementary Tables 1,2,6)
- Source membership: ucc-research-protocol-sctab-seed: Part of scTab published annotation evaluation (Supplementary Tables 1,2,6)
- UCE frozen embeddings + classifier — scTab Table 1: facets.method_types: Facets: method types ["foundation_model"] (scTab Methods: UCE embeddings with a linear probe; cell-type-20261009-config-wu2025-uce-onclass is foundation_model)
- UCE frozen embeddings + classifier: description: Description Linear probe on Universal Cell Embeddings: a logistic-regression cell-type classifier is fitted to the precomputed frozen UCE embeddings hosted by CELLxGENE. (ucc-research-source-sctab-paper: Methods/Universal Cell Embedding (UCE) (zero-shot))
- XGBoost — scTab Table 1: facets.method_types: Facets: method types ["supervised_machine_learning"] (scTab Methods: each classifier is trained on labelled cells)
- XGBoost: description: Description Gradient-boosted tree classifier for tabular data, used as a nonlinear cell-type annotation baseline. (ucc-research-source-sctab-paper: Results (nonlinear models benchmarked against the linear baseline))
Models, benchmarks and other records
Methods (6)
Configurations (11)
- CellTypist — scTab Table 1
- celltypist default parameters — scTab Table 6
- CIForm — scTab Table 1
- MLP — scTab Table 1
- Optimized linear logistic regression — scTab Table 1
- scGPT fine-tuned — scTab Table 1
- scGPT frozen embeddings + logistic regression — scTab Table 1
- scTab — scTab Table 1
- UCE frozen embeddings + classifier — scTab Table 1
- XGBoost — scTab Table 1
- xgboost default parameters — scTab Table 6
Benchmarks (1)
Protocols (4)
Datasets (1)
Baselines (3)
Evaluations (11)
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
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.
19 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Artifact SHA-256 e61b11cd873f3d13e891e9c33bac0847070826200f8fbe0b3969d994f1c52290 Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Artifact url https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11298532/fullTextXML Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Retrieval method HTTPS download; article XML via Europe PMC primary full-text mirror Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Retrieved at 2026-09-30T21:23:28Z Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Review method1 values
| scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Review note Primary-source table transcription and scope review; no independent reproduction or human scientific review. Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Review date 2026-09-30T21:43:56Z Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Reviewer note role: research coverage worker Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Reviewer1 values
| scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Source location Results; Methods; data and code availability Source metadata | scTab: Scaling cross-tissue single-cell annotation models No field-specific location recorded Version: 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30 | catalogued No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-11-21-c8bea25e5abb · Record review: source checked
Primary research source retrieved and inspected for this coverage audit.
Technical metadata and extraction receipts
Stable ID: ucc-research-source-sctab-paper
- url
- https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11298532/fullTextXML
- version
- 10.1038/s41467-024-51059-5; published article XML retrieved 2026-09-30
- retrieved at
- 2026-09-30T21:23:28Z
- artifact url
- https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11298532/fullTextXML
- source locator
- Results; Methods; data and code availability
- review
- method: automated-source-review; reviewed at: 2026-09-30T21:43:56Z; note: Primary-source table transcription and scope review; no independent reproduction or human scientific review.; reviewer note: role: research coverage worker; reviewer: codex
- retrieval method
- HTTPS download; article XML via Europe PMC primary full-text mirror
- artifact sha256
- e61b11cd873f3d13e891e9c33bac0847070826200f8fbe0b3969d994f1c52290