Task

E. coli sigma70 promoter prediction

The E. coli promoter task is an independent test setting within ProkBERT’s broader prokaryotic evaluation.

Sources

ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141

15 evaluations · 60 metric rows

At 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.

Data, procedure and scoring
Property	Description and evidence
Datasets	PPD-derived promoter data and a separately identified E. coli sigma70 test collection. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Splits	Independent E. coli testing is described; exact training exclusions remain unextracted. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Metrics	Sensitivity, specificity and accuracy are discussed together to expose false-positive tradeoffs. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Baselines	CNNProm, Sigma70Pred, iPromoter-BnCNN, iPromoter-2L and Promotech. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Leakage controls	The source warns that some comparator training data may overlap or be closely related to the E. coli evaluation data. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
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 sources Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Entity type	Paper-specific computational evaluation protocol. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Organisms	Escherichia coli for the sigma70 test. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Assays	PPD and independent sigma70 promoter annotations. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Allowed inputs	DNA promoter-window sequence. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141
Adaptation	Supervised promoter prediction compared with established task-specific methods. Sources ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141

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

ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141

Evaluation methodology

PPD-derived promoter data and a separately identified E. coli sigma70 test collection. Independent E. coli testing is described; exact training exclusions remain unextracted. Sensitivity, specificity and accuracy are discussed together to expose false-positive tradeoffs. CNNProm, Sigma70Pred, iPromoter-BnCNN, iPromoter-2L and Promotech. The source warns that some comparator training data may overlap or be closely related to the E. coli evaluation data. 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.

Sources

ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141

Evaluation design

Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.

Protocols

E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)

These source-backed links do not make different protocols or scores interchangeable.

Published comparisons

Explore the results reported under one evaluation protocol. Each figure keeps its source, dataset and metric together; it is not a ranking across studies.

Comparison and metric

E. coli sigma70 independent promoter test · Table 3

Accuracy (unitless) · Higher values are better for this metric.

Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training.

Evaluation protocol · E. coli sigma70 promoter dataset

ProkBERT-mini · Configuration · Author-reported evaluation0.87
ProkBERT-mini-c · Configuration · Author-reported evaluation0.87
ProkBERT-mini-long · Configuration · Author-reported evaluation0.87
CNNProm · Configuration · Result quoted from another source0.72
iPro70-FMWin · Configuration · Result quoted from another source0.76
70ProPred · Configuration · Result quoted from another source0.74
iPromoter-2L · Configuration · Result quoted from another source0.64
Multiply · Configuration · Result quoted from another source0.50
bTSSfinder · Configuration · Result quoted from another source0.46
BPROM · Configuration · Result quoted from another source0.56
IBPP · Configuration · Result quoted from another source0.50
Promotech · Configuration · Independent external evaluation0.71
Sigma70Pred · Configuration · Independent external evaluation0.66
iPromoter-BnCNN · Configuration · Independent external evaluation0.55
MULTiPly · Configuration · Independent external evaluation0.54

Source order is preserved. Plotted marks show point estimates; uncertainty, where reported, is retained in the printed values and table. Differences do not establish statistical significance.

ProkBERT family: genomic language models for microbiome applications · Table 3: Accuracy, E. coli sigma70 independent promoter test

Values, uncertainty and evidence

Accuracy: original source values
Tested entity	Printed value	Uncertainty	Evidence
ProkBERT-mini · Configuration	0.87 unitless	Not reported	Author-reported evaluation · source checkedProkBERT family: genomic language models for microbiome applications; ProkBERT family: genomic language models for microbiome applications · Table 3, ProkBERT-mini row, Accuracy column
ProkBERT-mini-c · Configuration	0.87 unitless	Not reported	Author-reported evaluation · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-c, column Accuracy; XML row3 column2
ProkBERT-mini-long · Configuration	0.87 unitless	Not reported	Author-reported evaluation · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-long, column Accuracy; XML row4 column2
CNNProm · Configuration	0.72 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row CNNProm, column Accuracy; XML row5 column2
iPro70-FMWin · Configuration	0.76 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPro70-FMWin, column Accuracy; XML row6 column2
70ProPred · Configuration	0.74 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row 70ProPred, column Accuracy; XML row7 column2
iPromoter-2L · Configuration	0.64 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-2L, column Accuracy; XML row8 column2
Multiply · Configuration	0.50 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Multiply, column Accuracy; XML row9 column2
bTSSfinder · Configuration	0.46 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row bTSSfinder, column Accuracy; XML row10 column2
BPROM · Configuration	0.56 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row BPROM, column Accuracy; XML row11 column2
IBPP · Configuration	0.50 unitless	Not reported	Result quoted from another source · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row IBPP, column Accuracy; XML row12 column2
Promotech · Configuration	0.71 unitless	Not reported	Independent external evaluation · source checkedProkBERT family: genomic language models for microbiome applications; ProkBERT family: genomic language models for microbiome applications · Table 3, Promotech row, Accuracy column
Sigma70Pred · Configuration	0.66 unitless	Not reported	Independent external evaluation · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Sigma70Pred, column Accuracy; XML row14 column2
iPromoter-BnCNN · Configuration	0.55 unitless	Not reported	Independent external evaluation · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-BnCNN, column Accuracy; XML row15 column2
MULTiPly · Configuration	0.54 unitless	Not reported	Independent external evaluation · source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row MULTiPly, column Accuracy; XML row16 column2

Scope and limitations

Historical comparator implementations and training sets differ; source table alone does not prove matched training budget.
No interval assigned unless printed in source cell.

Source transcription and grouping reviewed by automated source review on 2026-09-17. These experiments were not independently reproduced by rewire.

Tested entities and results

Release 2026-09-17-d277315f7d76 · 15 evaluations · 60 metric rows. Different protocols are not a single leaderboard.

Evaluation setupMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
ProkBERT-mini: E. coli sigma70 promoter prediction Configuration: ProkBERT-miniProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Author-reported evaluation · Evaluation metadata: needs review
0.87 Accuracy Unit: unitless · Direction: higher	Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications; ProkBERT family: genomic language models for microbiome applications · Table 3, ProkBERT-mini row, Accuracy column Source checking is not independent reproduction.
0.90 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini, column Sensitivity; XML row2 column4 Source checking is not independent reproduction.
0.85 Specificity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini, column Specificity; XML row2 column5 Source checking is not independent reproduction.
Promotech: E. coli sigma70 promoter prediction Configuration: PromotechProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Independent external evaluation · Evaluation metadata: needs review
0.71 Accuracy Unit: unitless · Direction: higher	Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications; ProkBERT family: genomic language models for microbiome applications · Table 3, Promotech row, Accuracy column Source checking is not independent reproduction.
0.49 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Promotech, column Sensitivity; XML row13 column4 Source checking is not independent reproduction.
Sigma70Pred: E. coli sigma70 independent promoter test Configuration: Sigma70PredProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Independent external evaluation · Evaluation metadata: needs review
0.66 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Sigma70Pred, column Accuracy; XML row14 column2 Source checking is not independent reproduction.
ProkBERT-mini-long: E. coli sigma70 independent promoter test Configuration: ProkBERT-mini-longProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Author-reported evaluation · Evaluation metadata: needs review
0.89 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-long, column Sensitivity; XML row4 column4 Source checking is not independent reproduction.
0.87 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-long, column Accuracy; XML row4 column2 Source checking is not independent reproduction.
CNNProm: E. coli sigma70 independent promoter test Configuration: CNNPromProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.51 Specificity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row CNNProm, column Specificity; XML row5 column5 Source checking is not independent reproduction.
iPromoter-2L: E. coli sigma70 independent promoter test Configuration: iPromoter-2LProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.64 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-2L, column Accuracy; XML row8 column2 Source checking is not independent reproduction.
0.37 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-2L, column MCC; XML row8 column3 Source checking is not independent reproduction.
0.37 Specificity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-2L, column Specificity; XML row8 column5 Source checking is not independent reproduction.
70ProPred: E. coli sigma70 independent promoter test Configuration: 70ProPredProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.74 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row 70ProPred, column Accuracy; XML row7 column2 Source checking is not independent reproduction.
0.51 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row 70ProPred, column MCC; XML row7 column3 Source checking is not independent reproduction.
bTSSfinder: E. coli sigma70 independent promoter test Configuration: bTSSfinderProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.46 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row bTSSfinder, column Accuracy; XML row10 column2 Source checking is not independent reproduction.
BPROM: E. coli sigma70 independent promoter test Configuration: BPROMProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.10 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row BPROM, column MCC; XML row11 column3 Source checking is not independent reproduction.
iPromoter-BnCNN: E. coli sigma70 independent promoter test Configuration: iPromoter-BnCNNProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Independent external evaluation · Evaluation metadata: needs review
0.18 Specificity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPromoter-BnCNN, column Specificity; XML row15 column5 Source checking is not independent reproduction.
IBPP: E. coli sigma70 independent promoter test Configuration: IBPPProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
-0.03 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row IBPP, column MCC; XML row12 column3 Source checking is not independent reproduction.
ProkBERT-mini-c: E. coli sigma70 independent promoter test Configuration: ProkBERT-mini-cProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Author-reported evaluation · Evaluation metadata: needs review
0.88 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-c, column Sensitivity; XML row3 column4 Source checking is not independent reproduction.
0.85 Specificity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row ProkBERT-mini-c, column Specificity; XML row3 column5 Source checking is not independent reproduction.
iPro70-FMWin: E. coli sigma70 independent promoter test Configuration: iPro70-FMWinProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.76 Accuracy Unit: unitless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPro70-FMWin, column Accuracy; XML row6 column2 Source checking is not independent reproduction.
0.53 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row iPro70-FMWin, column MCC; XML row6 column3 Source checking is not independent reproduction.
MULTiPly: E. coli sigma70 independent promoter test Configuration: MULTiPlyProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Independent external evaluation · Evaluation metadata: needs review
0.92 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row MULTiPly, column Sensitivity; XML row16 column4 Source checking is not independent reproduction.
Multiply: E. coli sigma70 independent promoter test Configuration: MultiplyProtocol: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)Dataset: E. coli sigma70 promoter dataset Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training. Result quoted from another source · Evaluation metadata: needs review
0.05 MCC Unit: dimensionless · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Multiply, column MCC; XML row9 column3 Source checking is not independent reproduction.
0.81 Sensitivity Unit: fraction · Direction: higher	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedProkBERT family: genomic language models for microbiome applications · Table 3, row Multiply, column Sensitivity; XML row9 column4 Source checking is not 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
ProkBERT family: genomic language models for microbiome applications	PMC10810988.1	Read source DOI: 10.3389/fmicb.2023.1331233

What is still missing

Independent batch review before import; preserve existing observation identities.

Search and extraction details

complete comparison tables extracted pending publication review

Searches

ProkBERT family: genomic language models for microbiome applications primary paper benchmark results

Evidence locations

Table3; independent E.coli sigma70 set in Section2.3

Strengths and limitations

Strengths and considerations

No source-reviewed explanatory claims are recorded here yet.

Limitations and conditions

Comparator source versions remain unextracted.
Sources
ProkBERT family: genomic language models for microbiome applications · Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141

Profile review details

Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced.

Stable record: reported-task-3891811dcce8b3

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.

17 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-09-17-d277315f7d76
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	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram steps ["Input: DNA promoter-window sequence.","Evaluation: Supervised promoter prediction compared with established task-specific methods.","Readout: Sensitivity, specificity and accuracy are discussed together to expose false-positive tradeoffs."] Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram title Computational evaluation flow Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Datasets PPD-derived promoter data and a separately identified E. coli sigma70 test collection. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Splits Independent E. coli testing is described; exact training exclusions remain unextracted. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Adaptation Supervised promoter prediction compared with established task-specific methods. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.10.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Metrics Sensitivity, specificity and accuracy are discussed together to expose false-positive tradeoffs. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.2.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Baselines CNNProm, Sigma70Pred, iPromoter-BnCNN, iPromoter-2L and Promotech. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.3.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Leakage controls The source warns that some comparator training data may overlap or be closely related to the E. coli evaluation data. Individual claims	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	source checked automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.4.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
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	ProkBERT family: genomic language models for microbiome applications Original source ↗ Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141 Version: PMC10810988.1 Retrieved: 2026-09-16T10:33:36.197Z	unreported automated source review · 2026-09-16 Audit details Task-specific computational methodology and field context checked in the cited primary-source artifact. Source-backed fields, inapplicable evaluator dimensions and unresolved details are distinguished. Numerical results were not reproduced. Field: `attributes.profile.facts.5.value` Source artifact SHA-256: `8610e2a54aa877c8dc565a9cdb6e82099f284c5e0907a52cab18d994ea732436` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact

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

Sources and history

Release 2026-09-17-d277315f7d76 · Record review: needs review

2 source records and release history

ProkBERT family: genomic language models for microbiome applications · Original source · PMC10810988.1
ProkBERT family: genomic language models for microbiome applications · Original source · PMC10810988.1

Download this release

Technical metadata and extraction receipts

Stable ID: reported-task-3891811dcce8b3

areas: microbes-communities
tasks: E. coli sigma70 promoter prediction
entity level: task
version: Not reported
task: E. coli sigma70 promoter prediction
scope note: Paper-specific evaluation task; protocol completeness requires further extraction.
comparison panels: id: part2-prokbert-2024-T3-28db1fe651; title: E. coli sigma70 independent promoter test · Table 3; protocol id: paper-protocol-94fa3494d41bcd98d2; dataset id: reported-dataset-48def1da574597; metric: Accuracy; unit: unitless; direction: higher; result ids: lit-033; paper-result-feb6117d798c298b3a; paper-result-4c8aea33e9c70cd11f; paper-result-6f98ff1a4772f9ccb3; paper-result-410891a647e6249a90; paper-result-13fec9cc9a02b9ade1; paper-result-0c5fed0bf3be3b8d58; paper-result-bc336af932cc34bfff; paper-result-1e1456e63f197bf736; paper-result-9b1c2838a500cd6be1; paper-result-68a650408dc07175aa; lit-034; paper-result-01c368320e32c0ab0e; paper-result-9656f7e500dc93dcbb; paper-result-8f3993306e0616cbfc; source ids: part2-prokbert-2024; source locator: Table 3: Accuracy, E. coli sigma70 independent promoter test; context: Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training.; caveats: Historical comparator implementations and training sets differ; source table alone does not prove matched training budget.; No interval assigned unless printed in source cell.; review: method: automated_source_review; date: 2026-09-17; id: part2-prokbert-2024-T3-8c8321b937; title: E. coli sigma70 independent promoter test · Table 3; protocol id: paper-protocol-94fa3494d41bcd98d2; dataset id: reported-dataset-48def1da574597; metric: MCC; unit: dimensionless; direction: higher; result ids: paper-result-feb8e65e7cf4e58b24; paper-result-60e6848b40696ac12e; paper-result-dc6aa1769ce2cd3f4e; paper-result-7a77ab58dafad4a2aa; paper-result-4735c3c4f977af0fce; paper-result-37c1083392f34f9406; paper-result-2c8339165057825390; paper-result-4a2c32c0125b04d6d4; paper-result-e75353105223f1e447; paper-result-208a44d26d306cd65a; paper-result-2a20fdc870d161095a; paper-result-5b13f4aa84f244913e; paper-result-a48e2ed7a547d27527; paper-result-4d73754c1cb397329c; paper-result-e8fc9cd21d61cf7e9d; source ids: part2-prokbert-2024; source locator: Table 3: MCC, E. coli sigma70 independent promoter test; context: Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training.; caveats: Historical comparator implementations and training sets differ; source table alone does not prove matched training budget.; No interval assigned unless printed in source cell.; review: method: automated_source_review; date: 2026-09-17; id: part2-prokbert-2024-T3-9956d5103b; title: E. coli sigma70 independent promoter test · Table 3; protocol id: paper-protocol-94fa3494d41bcd98d2; dataset id: reported-dataset-48def1da574597; metric: Sensitivity; unit: fraction; direction: higher; result ids: paper-result-26d268756b8a662fd8; paper-result-3d5f92c4cb77e67f50; paper-result-03b41950c492f3431d; paper-result-e555e071d181460c12; paper-result-6e84f7880c063cbdd7; paper-result-e29c79b8c8cd56be14; paper-result-b3b697a5be7a34fe24; paper-result-4ab4d0f27fcfd2347b; paper-result-ba12a8ea88fda90346; paper-result-7a7d041d187d9e7189; paper-result-d562aac34ec0538c9b; paper-result-074f43f6d5f3283258; paper-result-77a6d05afe1045f61a; paper-result-a560ff80bc0c25f435; paper-result-42d188ba97fceb115b; source ids: part2-prokbert-2024; source locator: Table 3: Sensitivity, E. coli sigma70 independent promoter test; context: Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training.; caveats: Historical comparator implementations and training sets differ; source table alone does not prove matched training budget.; No interval assigned unless printed in source cell.; review: method: automated_source_review; date: 2026-09-17; id: part2-prokbert-2024-T3-0f5f4b674d; title: E. coli sigma70 independent promoter test · Table 3; protocol id: paper-protocol-94fa3494d41bcd98d2; dataset id: reported-dataset-48def1da574597; metric: Specificity; unit: fraction; direction: higher; result ids: paper-result-2ad5c1987a41ae98d8; paper-result-43e98ef9fed2b6cabd; paper-result-c505c37d3a9bd19568; paper-result-0ba7173d41197d3830; paper-result-d6f5d354ed1d3656ba; paper-result-dc1256b9ef86c3feef; paper-result-442110ff6e4a09bb12; paper-result-e3fe9457bfda3bc5e9; paper-result-eb1bf95d87961d46a3; paper-result-6f3b1fc0ec99ea98ae; paper-result-65020951c24bd83246; paper-result-c9b2fec5bfd04f7818; paper-result-5927468ae054290ab3; paper-result-26e19e7a3c593c8e0d; paper-result-d7f6e83cdf5b05dd97; source ids: part2-prokbert-2024; source locator: Table 3: Specificity, E. coli sigma70 independent promoter test; context: Test-only evaluation on Cassiano and Silva-Rocha 2020 data; methods have different training histories. 865 high-evidence RegulonDB 10.5 promoters and 1,000 nucleotide-distribution-matched negative sequences. Promoter exact matches removed from model training.; caveats: Historical comparator implementations and training sets differ; source table alone does not prove matched training budget.; No interval assigned unless printed in source cell.; review: method: automated_source_review; date: 2026-09-17
benchmark research: review date: 2026-09-17; status: complete_comparison_tables_extracted_pending_publication_review; primary sources: part2-prokbert-2024; inspected locators: Table3; independent E.coli sigma70 set in Section2.3; searched queries: ProkBERT family: genomic language models for microbiome applications primary paper benchmark results; gaps: Independent batch review before import; preserve existing observation identities.; 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: 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: prokbert-2024; source locator: Methods §2.3.1; Results: promoter comparison; cached text lines 66–67, 138–141; 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: E. coli sigma70 promoter dataset
evaluates task: E. coli sigma70 independent promoter test (E. coli sigma70 promoter prediction)

Suggest a correction