Model

ChromBPNet

ChromBPNet predicts base-resolution chromatin accessibility while modeling assay-specific enzyme bias separately.

Sources (2)

kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params

13 evaluations · 17 results · 1 evaluated configuration using this model

How it worksChromBPNet workflow

Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation.

Sources (2)

Overview

Model type

Bias-factorized convolutional chromatin-profile predictor

Inputs

DNA sequences and ATAC-seq or DNase-seq profiles for the configured assay.

Outputs

Predicted accessibility profiles and quantities used to study sequence contributions and variants.

Access

Official project documentation and implementation: https://github.com/kundajelab/chrombpnet

Sources (2)

limited source coverage · Automated source review, 2026-09-16. All specifications and missing details

Evaluations and results

13 evaluations · 17 results. Different protocols are not a single leaderboard.

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: chrombpnet (paper Table 4)	Protocol: SPI1 binding QTL classification (AlphaGenome paper) Dataset subset: SPI1 binding QTL classification: evaluated data subset	0.356 auPRC dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): SPI1 binding QTL classification Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: auPRC over causal/noncausal labels. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L36
Configuration: ChromBPNet (paper Table 3)	Protocol: ATAC prediction on held-out peaks (AlphaGenome paper) Dataset subset: ATAC prediction on held-out peaks: evaluated data subset	0.467 profile jsd dimensionless · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): ATAC prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J30
Configuration: chrombpnet (paper Table 4)	Protocol: European-ancestry LCL caQTL classification (AlphaGenome paper) Dataset subset: European-ancestry LCL caQTL classification: evaluated data subset	0.279 auPRC dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): European-ancestry LCL caQTL classification Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: auPRC over causal/noncausal labels. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L22
Configuration: chrombpnet (paper Table 4)	Protocol: Zero-shot CAGI5 MPRA activity-effect prediction (chrombpnet-matched comparison) (AlphaGenome paper) Dataset subset: Zero-shot CAGI5 MPRA activity-effect prediction (chrombpnet-matched comparison): evaluated data subset	0.544 mean_pearsonr_all correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): Zero-shot CAGI5 MPRA activity-effect prediction (chrombpnet-matched comparison) Compute locus/context Pearson correlations between predicted and observed effects. The Borzoi strategy uses its reported scoring windows and modified matching; the ChromBPNet comparison excludes the unavailable TERT-GBM context. Aggregation: Mean Pearson correlation over the included locus/context comparisons; Table4 rows10 and11 have different AlphaGenome scalars and must remain separate. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L10
Configuration: chrombpnet (paper Table 4)	Protocol: African-ancestry LCL caQTL classification (AlphaGenome paper) Dataset subset: African-ancestry LCL caQTL classification: evaluated data subset	0.415 auPRC dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): African-ancestry LCL caQTL classification Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: auPRC over causal/noncausal labels. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L20
Configuration: chrombpnet (paper Table 4)	Protocol: Coronary smooth-muscle caQTL effect-size prediction (AlphaGenome paper) Dataset subset: Coronary smooth-muscle caQTL effect-size prediction: evaluated data subset	0.658 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): Coronary smooth-muscle caQTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L34
Configuration: chrombpnet (paper Table 4)	Protocol: SPI1 binding QTL effect-size prediction (AlphaGenome paper) Dataset subset: SPI1 binding QTL effect-size prediction: evaluated data subset	0.52 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): SPI1 binding QTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L38
Configuration: chrombpnet (paper Table 4)	Protocol: Microglia caQTL effect-size prediction (AlphaGenome paper) Dataset subset: Microglia caQTL effect-size prediction: evaluated data subset	0.61 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): Microglia caQTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L32
Configuration: ChromBPNet (paper Table 3)	Protocol: ATAC prediction on held-out peaks (AlphaGenome paper) Dataset subset: ATAC prediction on held-out peaks: evaluated data subset	0.78 log1p_count_pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): ATAC prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J29
Configuration: ChromBPNet (paper Table 3)	Protocol: DNase prediction on held-out peaks (AlphaGenome paper) Dataset subset: DNase prediction on held-out peaks: evaluated data subset	0.66 log1p_count_pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): DNase prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J23
Configuration: chrombpnet (paper Table 4)	Protocol: Yoruba LCL dsQTL classification (AlphaGenome paper) Dataset subset: Yoruba LCL dsQTL classification: evaluated data subset	0.543 auPRC dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): Yoruba LCL dsQTL classification Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: auPRC over causal/noncausal labels. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L24
Configuration: chrombpnet (paper Table 4)	Protocol: African-ancestry LCL caQTL effect-size prediction (AlphaGenome paper) Dataset subset: African-ancestry LCL caQTL effect-size prediction: evaluated data subset	0.674 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): African-ancestry LCL caQTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L26
Configuration: ChromBPNet (paper Table 3)	Protocol: ATAC prediction on held-out peaks (AlphaGenome paper) Dataset subset: ATAC prediction on held-out peaks: evaluated data subset	0.786 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): ATAC prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J28
Configuration: ChromBPNet (paper Table 3)	Protocol: DNase prediction on held-out peaks (AlphaGenome paper) Dataset subset: DNase prediction on held-out peaks: evaluated data subset	0.527 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): DNase prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J22
Configuration: chrombpnet (paper Table 4)	Protocol: Yoruba LCL dsQTL effect-size prediction (AlphaGenome paper) Dataset subset: Yoruba LCL dsQTL effect-size prediction: evaluated data subset	0.772 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): Yoruba LCL dsQTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L30
Configuration: ChromBPNet (paper Table 3)	Protocol: DNase prediction on held-out peaks (AlphaGenome paper) Dataset subset: DNase prediction on held-out peaks: evaluated data subset	0.561 profile_jsd dimensionless · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChromBPNet (paper Table 3): DNase prediction on held-out peaks Evaluate signal correlation, log-total-count correlation and profile JSD as separate metrics; retain only endpoints actually listed for this evaluation in Table3. Aggregation: Table3 reports task summary scalars; Extended Data Fig.3 shows per-cell-line comparisons. The inspected caption does not fully specify pooling across peaks/cell lines. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 3 Track performance'!J24
Configuration: chrombpnet (paper Table 4)	Protocol: European-ancestry LCL caQTL effect-size prediction (AlphaGenome paper) Dataset subset: European-ancestry LCL caQTL effect-size prediction: evaluated data subset	0.525 pearsonr correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source chrombpnet (paper Table 4): European-ancestry LCL caQTL effect-size prediction Use the source’s local log-fold-change scorer, select the task-specific cell/assay tracks and compare against the source-provided QTL labels or measured effects. Aggregation: Pearson correlation against reported effect sizes of the causal/significant QTL set. AlphaGenome Nature 2026 supplementary comparison tables · 'Suppl Table 4 Variant performan'!L28

Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.

Related configurations, pipelines and services

These configurations, services and pipelines use this model within their own configurations. Their results, where available, are not assigned to the underlying model.

ARSENAL+ChromBPNet Pipeline · 1 result

Use this model

How it works, versions and access

Versions and evaluated configurations

ChromBPNet (paper Table 3) · family
chrombpnet (paper Table 4) · family

How it works

ChromBPNet predicts base-resolution chromatin accessibility while modeling assay-specific enzyme bias separately. Residual dilated convolutional network with a frozen bias model and a transcription-factor sequence component. The documented inputs are DNA sequences and ATAC-seq or DNase-seq profiles for the configured assay. The output consists of predicted accessibility profiles and quantities used to study sequence contributions and variants.

Sources (2)

Versions and reproducibility

Version-dependent trained models; README highlights a motif-discovery note for versions at or below 0.1.3. inputlen and outputlen are explicit model configuration fields. Record both the input sequence window and the central output window for the selected trained model.

Sources (2)

Strengths, limitations and unresolved questions

Strengths and limitations

Strengths and considerations

The two-stage procedure explicitly separates a learned assay-bias component from the sequence component of interest.
Sources (2)
kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params

Limitations and conditions

The bias model must correspond to the assay. The model is fitted to a specified context, so a checkpoint cannot be assumed to apply equally across assays or cell types.
Sources (2)
kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params

Profile review details

Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied.

Stable record: discovery-model-chrombpnet

Specifications

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.

Inputs, outputs and configuration
Property	Description and evidence
Model type	Bias-factorized convolutional chromatin-profile predictor Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Architecture	Residual dilated convolutional network with a frozen bias model and a transcription-factor sequence component. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Inputs	DNA sequences and ATAC-seq or DNase-seq profiles for the configured assay. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Outputs	Predicted accessibility profiles and quantities used to study sequence contributions and variants. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Parameters	Configuration-dependent: convolutional filter count and number of dilated layers are supplied in model_params; the bias component is separate. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Known versions	Version-dependent trained models; README highlights a motif-discovery note for versions at or below 0.1.3. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Training data	Two-stage fitting: background regions for enzyme bias, then accessibility-profile training for the sequence model. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Training cutoff	Inapplicable as a universal pretraining date: the workflow fits the supplied chromatin tracks and bias model; their accessions, dates and split belong to the individual evaluation. · Not applicable Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Context limits	inputlen and outputlen are explicit model configuration fields. Record both the input sequence window and the central output window for the selected trained model. Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Weights licence	Separate checkpoint-distribution terms are not stated in the inspected release documentation and licence material. The source-code licence alone is not recorded as an explicit weight grant. · Not reported in inspected sources Sources (3) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py; kundajelab/chrombpnet: LICENSE · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params; LICENSE: licence text
Access	Official project documentation and implementation: https://github.com/kundajelab/chrombpnet Sources (2) kundajelab/chrombpnet: README.md; kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params
Code licence	MIT Sources kundajelab/chrombpnet: LICENSE · LICENSE: licence text

Applicable tests and references

Applicability is distinct from a completed evaluation.

DART-Eval · Proposed association

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.

40 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-09-29-06401fd5b220
Property and statement	Original source and location	Review and provenance
Diagram caption Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation. Individual claims	kundajelab/chrombpnet: README.md Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `058c4f98218ea2ed854681126b6f682c9f3beec91275781fb37e39c55d0c92ea` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Diagram caption Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation. Individual claims	kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `bedc63a36ee27bb25ab39a00f1d9e3d4d64809754bf46aebe10d17b3a331dd33` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Diagram steps Background assay data Learn enzyme-bias model Fit sequence model with frozen bias Accessibility predictions Individual claims	kundajelab/chrombpnet: README.md Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `058c4f98218ea2ed854681126b6f682c9f3beec91275781fb37e39c55d0c92ea` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Diagram steps Background assay data Learn enzyme-bias model Fit sequence model with frozen bias Accessibility predictions Individual claims	kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `bedc63a36ee27bb25ab39a00f1d9e3d4d64809754bf46aebe10d17b3a331dd33` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Diagram title ChromBPNet workflow Individual claims	kundajelab/chrombpnet: README.md Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `058c4f98218ea2ed854681126b6f682c9f3beec91275781fb37e39c55d0c92ea` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Diagram title ChromBPNet workflow Individual claims	kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `bedc63a36ee27bb25ab39a00f1d9e3d4d64809754bf46aebe10d17b3a331dd33` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Model type Bias-factorized convolutional chromatin-profile predictor Individual claims	kundajelab/chrombpnet: README.md Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `058c4f98218ea2ed854681126b6f682c9f3beec91275781fb37e39c55d0c92ea` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Model type Bias-factorized convolutional chromatin-profile predictor Individual claims	kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `bedc63a36ee27bb25ab39a00f1d9e3d4d64809754bf46aebe10d17b3a331dd33` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Architecture Residual dilated convolutional network with a frozen bias model and a transcription-factor sequence component. Individual claims	kundajelab/chrombpnet: README.md Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `058c4f98218ea2ed854681126b6f682c9f3beec91275781fb37e39c55d0c92ea` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact
Architecture Residual dilated convolutional network with a frozen bias model and a transcription-factor sequence component. Individual claims	kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py Original source ↗ README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: 09938fdb4397ec0006510e5251e48920a505d4de Retrieved: 2026-09-16T19:46:19.514840+00:00	source checked automated source review · 2026-09-16 Audit details Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `bedc63a36ee27bb25ab39a00f1d9e3d4d64809754bf46aebe10d17b3a331dd33` Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact Inspected artifact

All evidence (CSV)All evidence (JSONL)How to read the evidence

Sources and history

View linked audit checks and correction history

Release 2026-09-29-06401fd5b220 · Record review: discovered

4 source records and release history

kundajelab/chrombpnet official source · Original source · 09938fdb4397ec0006510e5251e48920a505d4de
kundajelab/chrombpnet: README.md · Original source · 09938fdb4397ec0006510e5251e48920a505d4de
kundajelab/chrombpnet: chrombpnet/training/models/chrombpnet_with_bias_model.py · Original source · 09938fdb4397ec0006510e5251e48920a505d4de
kundajelab/chrombpnet: LICENSE · Original source · 09938fdb4397ec0006510e5251e48920a505d4de

Download this release

Technical metadata and extraction receipts

Stable ID: discovery-model-chrombpnet

areas: genomics
access: official_source_linked
benchmark applicability: candidate; not evidence of a reported evaluation
candidate benchmark ids: discovery-benchmark-dart-eval
entity level: family
reported name: ChromBPNet
version: Not reported
historical missing metadata: checkpoint: unextracted; code licence: unextracted; parameters: unextracted; training cutoff: unextracted; training data: unextracted; version: unextracted; weights licence: 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 named learned biological predictor or representation model/family. Preserve this identity separately from task-specific fitting, individual checkpoints, pipelines and hosted access.; source ids: evidence-official-a766fb9176a491dc1ec3; evidence-official-bc3f1b0fa5b0999bbfa0; source locator: README.md: introductory explanation and Bias-factorized ChromBPNet training; chrombpnet/training/models/chrombpnet_with_bias_model.py: bpnet_model and model_params; ambiguities: None recorded

Related records

Suggest a correction