Strengths and considerations
No source-reviewed explanatory claims are recorded here yet.
Task
Cell-type structure in frozen embeddings
Frozen cell embeddings are examined for cell structure and patient-associated batch patterns.
Sources
GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61At a glance
Inputs, training, access and other details
Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.
| Property | Description and evidence |
|---|---|
| Datasets | The Aorta single-cell dataset is one analyzed collection, with patient and phenotype annotations.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Splits | The frozen-embedding clustering analysis does not define a supervised train/test split. Random holdout used for phenotype-classifier experiments belongs to a different evaluation. · Not reported in inspected sourcesSourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Metrics | Adjusted Rand index compares embedding-derived clusters with patient and phenotype labels.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Baselines | Geneformer, GenePT-s, scGPT and original expression measurements.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Leakage controls | The clustering comparison measures cell and patient associations in embeddings; it does not establish an independent unseen-patient supervised test. · Not reported in inspected sourcesSourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Uncertainty | The cited text-accessible evaluation sections give no confidence-interval, resampling or repeat-run error-bar specification. Image-only tables and uninspected supplements are outside this absence claim. · Not reported in inspected sourcesSourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Entity type | Paper-specific computational evaluation protocol.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Organisms | Human tissue cohorts are explicitly established for the aorta, arterial plaque, osteoarthritis cartilage and multiple-sclerosis brain datasets by their original studies. GenePT also evaluates pancreas and tumor-infiltrating myeloid collections. A complete organism-by-subset census is not supplied in the GenePT cell-task description.Sources (4)GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT; PMC9556750.xml; PMC7331607.xml; PMC6731122.xml · GenePT: Methods, Cell-level tasks and references 39–43; Alsaigh et al. Abstract; Chou et al. Results; Schirmer et al. Methods/Human tissue samples |
| Assays | Single-cell expression with cell-type, patient and phenotype annotations.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Allowed inputs | Gene-expression-derived cell embeddings and GenePT representations.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
| Adaptation | Frozen-embedding geometry assessment; no task-classifier result is implied by cell separation alone.SourcesGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 |
How it works
How it worksComputational evaluation flow
Conceptual summary of the cited evaluation; exact task configuration and source version remain part of the protocol.
Sources
GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61Evaluation methodology
The Aorta single-cell dataset is one analyzed collection, with patient and phenotype annotations. The frozen-embedding clustering analysis does not define a supervised train/test split. Random holdout used for phenotype-classifier experiments belongs to a different evaluation. Adjusted Rand index compares embedding-derived clusters with patient and phenotype labels. Geneformer, GenePT-s, scGPT and original expression measurements. The clustering comparison measures cell and patient associations in embeddings; it does not establish an independent unseen-patient supervised test.
Sources
GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Results: cell clustering and batch effects; Discussion; cached text lines 57, 61Recorded evaluations
Each evaluation records what was tested and under which conditions.
Tested entities and results
Release 2026-09-17-d277315f7d76 · 2 evaluations · 2 metric rows. Different protocols are not a single leaderboard.
| Metric and finding | Coverage and uncertainty | Evidence |
|---|---|---|
| GenePT-w: Cell-type structure in frozen embeddings Configuration: GenePT-wTask: Cell-type structure in frozen embeddingsDataset: Aorta single-cell dataset k-means on pretrained cell embeddings; agreement with original cell-type labels. Author-reported evaluation · Evaluation metadata: needs review | ||
| 0.54 Adjusted Rand Index Unit: unitless · Direction: unknown | Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported | source checkedGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Table 2, Aorta / Cell type row, GenePT-w ARI column Source checking is not independent reproduction. |
| scGPT: Cell-type structure in frozen embeddings Configuration: scGPTTask: Cell-type structure in frozen embeddingsDataset: Aorta single-cell dataset k-means on pretrained cell embeddings; agreement with original cell-type labels. Independent external evaluation · Evaluation metadata: needs review | ||
| 0.47 Adjusted Rand Index Unit: unitless · Direction: unknown | Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported | source checkedGenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Table 2, Aorta / Cell type row, scGPT ARI column Source checking is not independent reproduction. |
Papers and result coverage
Last literature check: 2026-09-17. Dated primary-source discovery and protocol/table screening. Source checking does not mean experimental reproduction. Only separately extracted and independently reviewed numeric batches are publishable.
| Paper or primary resource | Version | Reference |
|---|---|---|
| GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT | PMC archival version PMC10614824.2 | Read source |
What is still missing
- complete numerical transcription and independent cell review: Full primary artifact and table inventory preserved; no new numeric row is published from this audit alone.
- exact checkpoint hashes and per-method scored denominators: Table labels alone do not establish these fields; do not infer checkpoint or scored count from model name or dataset size.
Search and extraction details
primary comparison tables located
Searches
- GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT 10.1101/2023.10.16.562533
Evidence locations
- Table 2:; XML table T6
Strengths and limitations
Limitations and conditions
- Clustering agreement is not donor-held-out classification. The number of clusters is chosen to match annotated classes, and text-derived gene representations depend on the available descriptions.
Sources (4)
GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT; PMC9556750.xml; PMC7331607.xml; PMC6731122.xml · GenePT: Methods, Cell-level tasks and references 39–43; Alsaigh et al. Abstract; Chou et al. Results; Schirmer et al. Methods/Human tissue samples
Profile review details
Targeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change.
Stable record: reported-task-4df1fb456d3debEvidence 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.
23 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Diagram caption Conceptual summary of the cited evaluation; exact task configuration and source version remain part of the protocol. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram steps ["Input: Gene-expression-derived cell embeddings and GenePT representations.","Evaluation: The frozen-embedding clustering analysis does not define a supervised train/test split. Random holdout used for phenotype-classifier experiments belongs to a different evaluation.","Readout: Adjusted Rand index compares embedding-derived clusters with patient and phenotype labels."] Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram title Computational evaluation flow Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Datasets The Aorta single-cell dataset is one analyzed collection, with patient and phenotype annotations. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Splits The frozen-embedding clustering analysis does not define a supervised train/test split. Random holdout used for phenotype-classifier experiments belongs to a different evaluation. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | unreported automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Adaptation Frozen-embedding geometry assessment; no task-classifier result is implied by cell separation alone. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Metrics Adjusted Rand index compares embedding-derived clusters with patient and phenotype labels. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Baselines Geneformer, GenePT-s, scGPT and original expression measurements. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Leakage controls The clustering comparison measures cell and patient associations in embeddings; it does not establish an independent unseen-patient supervised test. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | unreported automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Uncertainty The cited text-accessible evaluation sections give no confidence-interval, resampling or repeat-run error-bar specification. Image-only tables and uninspected supplements are outside this absence claim. Individual claims | GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT Results: cell clustering and batch effects; Discussion; cached text lines 57, 61 Version: PMC archival version PMC10614824.2 | unreported automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-09-17-d277315f7d76 · Record review: needs review
5 source records and release history
- GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Original source · PMC archival version PMC10614824.2
- GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT · Original source · PMC archival version PMC10614824.2
- PMC9556750.xml · Original source · Retrieved 2026-09-16; sha256:71fd84bba280e9ac7b1009600432245122e73f6a01916a743ef4faeb56bb26f0
- PMC7331607.xml · Original source · Retrieved 2026-09-16; sha256:d87b7b7a67445c04db4f1fbb0cc52914b1bccd04a23d4ff6d66e2be5373291c6
- PMC6731122.xml · Original source · Retrieved 2026-09-16; sha256:2136d6c7589d2573e33c69f2b1eb0f2b76c696da06ee1c14427ccbaa45a33d5f
Technical metadata and extraction receipts
Stable ID: reported-task-4df1fb456d3deb
- areas
- cells-tissues
- tasks
- Cell-type structure in frozen embeddings
- entity level
- task
- version
- Not reported
- task
- Cell-type structure in frozen embeddings
- scope note
- Paper-specific evaluation task; protocol completeness requires further extraction.
- benchmark research
- review date: 2026-09-17; status: primary_comparison_tables_located; primary sources: evidence-expansion-genept-2024-230a2ec5; inspected locators: Table 2:; XML table T6; searched queries: GenePT: A Simple But Effective Foundation Model for Genes and Cells Built From ChatGPT 10.1101/2023.10.16.562533; gaps: complete numerical transcription and independent cell review: Full primary artifact and table inventory preserved; no new numeric row is published from this audit alone.; exact checkpoint hashes and per-method scored denominators: Table labels alone do not establish these fields; do not infer checkpoint or scored count from model name or dataset size.; claim scope: Dated primary-source discovery and protocol/table screening. Source checking does not mean experimental reproduction. Only separately extracted and independently reviewed numeric batches are publishable.
- historical missing metadata
- protocol version: not_reported_in_legacy_extract; split: not_reported_in_legacy_extract
- 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.
- legacy kinds
- benchmark
- entity classification
- review date: 2026-09-17; rationale: This source-scoped record identifies the biological prediction task and holds its paper context. Preserve the existing task identity; exact split, model adaptation and scoring remain in linked evaluations or separate protocol records.; source ids: genept-2024; source locator: Results: cell clustering and batch effects; Discussion; cached text lines 57, 61; ambiguities: A paper- or suite-specific task may constrain some inputs or metrics; that alone does not make it interchangeable with a complete versioned protocol. No protocol equivalence is inferred.; Some legacy profile Entity type facts use the generic phrase computational evaluation protocol. That boilerplate is not sufficient to establish a single fixed protocol identity or to merge this task with another protocol record.
Related records
- dataset: Aorta single-cell dataset
- benchmark: GenePT-w: Cell-type structure in frozen embeddings
- benchmark: scGPT: Cell-type structure in frozen embeddings