Task

RNA secondary structure

RNA secondary-structure prediction is evaluated separately for sequence-level and RNA-family generalization.

Sources

Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions

40 evaluations · 160 results

Overview

Datasets

ArchiveII, bpRNA-TS0, Rfam12.3–14.10 and experimentally grounded PDB50 evaluation collections.

Metrics

Macro-averaged precision, recall, F1 and interaction network fidelity on canonical base-pair predictions.

Allowed inputs

RNA sequence.

Sources

Evaluation procedure diagram

How it worksComputational evaluation flow

Conceptual summary of the cited evaluation; exact task configuration and source version remain part of the protocol.

Sources

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

Results

Each comparison retains its reviewed evaluation scope, dataset and metric. Results are shown without a pooled ranking.

Find a protocol, dataset or metric

Evaluation scope and datasetMetric and reported setup

bpRNA-TS0 · INF

INF (unitless) · Higher values are better.

bpRNA-TS0 (RNA secondary structure) · bpRNA-TS0

Evidence origin: Author-reported evaluation, Independent external evaluation.

Deep generalizable prediction of RNA secondary structure via base pair motif energy · Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 4 SPOT-RNA, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 5 MXfold2, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 6 ContextFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 7 CONTRAfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 8 EternaFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 9 LinearFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 10 RNAfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 11 SimFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 12 RNAstructure, column 2: bpRNA-TS0 INF

Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.

Comparison details and limitations

Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.

Automated source review: 2026-09-17. Numerical source review does not establish independent reproduction.

Dots show point estimates. Whiskers show only explicitly defined uncertainty (standard deviation, standard error or a labelled interval); their definitions remain in Table. Unresolved uncertainty is not plotted. Differences do not establish statistical significance.

Find a tested model or method

Showing 10 of 10 matching rows.

Zoom to observed range

Tested configuration

0.5110.5530.5950.6370.679

Reported score

Methods and evaluation design

Procedure, tasks and evaluated configurations

How it works

Evaluation methodology

ArchiveII, bpRNA-TS0, Rfam12.3–14.10 and experimentally grounded PDB50 evaluation collections. The paper distinguishes sequence-wise assessment from cross-family evaluation. Macro-averaged precision, recall, F1 and interaction network fidelity on canonical base-pair predictions. SPOT-RNA, MXfold2, ContextFold, CONTRAfold, EternaFold, LinearFold, RNAfold, SimFold and RNAstructure. A model confidence index is correlated with observed F1; that diagnostic is not a confidence interval for benchmark performance.

Sources

Evaluation design

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

Protocols

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

Run this benchmark

Choose a concrete protocol before running an evaluation. Its inputs, split and scoring rules determine which results can be compared.

Run instructions

No runnable recipe has been reviewed for this task. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

A task describes a biological question. Choose a linked protocol to obtain concrete split and scoring instructions.

Strengths, limitations and unresolved questions

Strengths and limitations

Strengths and considerations

No source-reviewed explanatory claims are recorded here yet.

Limitations and conditions

Sequence-wise and family-wise tests are distinct conditions. RNA-family exclusion is not a substitute for a full provenance audit of every training component and comparator.
Sources
Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions

Profile review details

Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged.

Stable record: reported-task-dc82fcbfb44935

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.

Data, procedure and scoring
Property	Description and evidence
Datasets	ArchiveII, bpRNA-TS0, Rfam12.3–14.10 and experimentally grounded PDB50 evaluation collections. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Splits	The paper distinguishes sequence-wise assessment from cross-family evaluation. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Metrics	Macro-averaged precision, recall, F1 and interaction network fidelity on canonical base-pair predictions. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Baselines	SPOT-RNA, MXfold2, ContextFold, CONTRAfold, EternaFold, LinearFold, RNAfold, SimFold and RNAstructure. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Leakage controls	The bpRNA benchmark applies an 80% sequence-similarity filter. Family-wise testing uses newly added Rfam families absent from the bpRNA training collection and separately removes similar sequences at 80%. Sequence-wise and unseen-family results therefore measure different forms of generalization. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Methods dataset list in full XML; Results: Evaluating BPfold on family-wise datasets; Tables 1–2
Uncertainty	A model confidence index is correlated with observed F1; that diagnostic is not a confidence interval for benchmark performance. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Entity type	Paper-specific computational evaluation protocol. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Organisms	The evaluation pools RNA-family datasets and PDB RNA structures. The Methods dataset list enumerates RNA families and sequence sets, not taxa. Individual bacterial and viral examples in Figure 6 do not establish the species composition of the aggregate benchmark. · Not reported in inspected sources Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Methods dataset list; Tables 1–2; Fig.6 caption
Assays	RNA secondary-structure references including experimentally grounded PDB structures. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Allowed inputs	RNA sequence. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions
Adaptation	Supervised sequence-to-structure prediction with separate sequence-wise and cross-family assessments. Sources Deep generalizable prediction of RNA secondary structure via base pair motif energy · Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Papers and result coverage

Last literature check: 2026-09-17. 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
Deep generalizable prediction of RNA secondary structure via base pair motif energy	version of record	Read source

Historical gaps recorded on 2026-09-17

The catalogue now holds 160 result rows for this benchmark. A note below about pending extraction describes the state on 2026-09-17 and may since have been answered by a later batch. The result rows and their sources are the current record.

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

complete tables extracted

Searches

Deep generalizable prediction of RNA secondary structure via base pair motif energy 10.1038/s41467-025-60048-1

Evidence locations

Table 1; XML table Tab1
Table 2; XML table Tab2

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 2026-09-29-06401fd5b220
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	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram steps Input: RNA sequence. Evaluation: Supervised sequence-to-structure prediction with separate sequence-wise and cross-family assessments. Readout: Macro-averaged precision, recall, F1 and interaction network fidelity on canonical base-pair predictions. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram title Computational evaluation flow Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Datasets ArchiveII, bpRNA-TS0, Rfam12.3–14.10 and experimentally grounded PDB50 evaluation collections. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Splits The paper distinguishes sequence-wise assessment from cross-family evaluation. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Adaptation Supervised sequence-to-structure prediction with separate sequence-wise and cross-family assessments. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.10.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Metrics Macro-averaged precision, recall, F1 and interaction network fidelity on canonical base-pair predictions. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.2.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Baselines SPOT-RNA, MXfold2, ContextFold, CONTRAfold, EternaFold, LinearFold, RNAfold, SimFold and RNAstructure. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.3.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Leakage controls The bpRNA benchmark applies an 80% sequence-similarity filter. Family-wise testing uses newly added Rfam families absent from the bpRNA training collection and separately removes similar sequences at 80%. Sequence-wise and unseen-family results therefore measure different forms of generalization. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Methods dataset list in full XML; Results: Evaluating BPfold on family-wise datasets; Tables 1–2 Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.4.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Uncertainty A model confidence index is correlated with observed F1; that diagnostic is not a confidence interval for benchmark performance. Individual claims	Deep generalizable prediction of RNA secondary structure via base pair motif energy Original source ↗ Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions Version: version of record Retrieved: 2026-09-16T10:41:16.502000+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.5.value` Source artifact SHA-256: `976218bd172998a1a6e7ed1609ecb8cb2ee380fb48a8dc7b25bc05ea8b0a49af` 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

View linked audit checks and correction history

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

2 source records and release history

Deep generalizable prediction of RNA secondary structure via base pair motif energy · Original source · version of record
Deep generalizable prediction of RNA secondary structure via base pair motif energy · Original source · version of record

Download this release

Technical metadata and extraction receipts

Stable ID: reported-task-dc82fcbfb44935

areas: rna-transcriptomes
tasks: RNA secondary structure
entity level: task
version: Not reported
task: RNA secondary structure
scope note: Paper-specific evaluation task; protocol completeness requires further extraction.
comparison panels: id: bpfold-2025-tab1-bprna-ts0-inf; title: bpRNA-TS0 · INF; protocol id: paper-protocol-208a8085432aa64eb2; dataset id: paper-dataset-9a595197e5581413b6; metric: INF; unit: unitless; direction: higher; result ids: paper-result-0458271bb982e75179; paper-result-5a1624fdb7640263f2; paper-result-7b916d75babf7a4d06; paper-result-e8e6954c70547446a1; paper-result-0da7bdbbf7b43c8456; paper-result-5249c6fca0adf30079; paper-result-5ebc9d5244115922e3; paper-result-66830706f75e525dbf; paper-result-711a4c604b04a9b0f2; paper-result-53887e7500c1cf009e; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 4 SPOT-RNA, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 5 MXfold2, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 6 ContextFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 7 CONTRAfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 8 EternaFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 9 LinearFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 10 RNAfold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 11 SimFold, column 2: bpRNA-TS0 INF; Table 1 (Tab1), row 12 RNAstructure, column 2: bpRNA-TS0 INF; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-bprna-ts0-f1; title: bpRNA-TS0 · F1; protocol id: paper-protocol-208a8085432aa64eb2; dataset id: paper-dataset-9a595197e5581413b6; metric: F1; unit: unitless; direction: higher; result ids: paper-result-f0f7111a6ec6426675; paper-result-4cd8df587d039cd7e1; paper-result-a5fe28b91e06263870; paper-result-8a436775a6db0626a1; paper-result-681cbde13cc79ef7ca; paper-result-0a1d1b57a38ce34e75; paper-result-35d6ca02cce090966b; paper-result-dcf0e23f78d1cb6fc0; paper-result-c536ed0e0b21471c4e; paper-result-1ccf4abe34808409ff; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 4 SPOT-RNA, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 5 MXfold2, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 6 ContextFold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 7 CONTRAfold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 8 EternaFold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 9 LinearFold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 10 RNAfold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 11 SimFold, column 3: bpRNA-TS0 F1; Table 1 (Tab1), row 12 RNAstructure, column 3: bpRNA-TS0 F1; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-bprna-ts0-precision; title: bpRNA-TS0 · Precision; protocol id: paper-protocol-208a8085432aa64eb2; dataset id: paper-dataset-9a595197e5581413b6; metric: Precision; unit: unitless; direction: higher; result ids: paper-result-11c1675cb5828808ab; paper-result-cb40146a919da9086a; paper-result-b78905513200ad10fe; paper-result-1069d4f8df9404d6e5; paper-result-deff3015ebe9868640; paper-result-5efbf1aebc52ee2283; paper-result-62e9fb6e2b7036a5d5; paper-result-ce1b188daa1f9f9fa6; paper-result-d7b4db6a7311486ac5; paper-result-bae3bdb96c31f1a771; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 4 SPOT-RNA, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 5 MXfold2, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 6 ContextFold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 7 CONTRAfold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 8 EternaFold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 9 LinearFold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 10 RNAfold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 11 SimFold, column 4: bpRNA-TS0 Precision; Table 1 (Tab1), row 12 RNAstructure, column 4: bpRNA-TS0 Precision; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-bprna-ts0-recall; title: bpRNA-TS0 · Recall; protocol id: paper-protocol-208a8085432aa64eb2; dataset id: paper-dataset-9a595197e5581413b6; metric: Recall; unit: unitless; direction: higher; result ids: paper-result-84076a2caea4d91a05; paper-result-362a0ed0c13044cdfc; paper-result-9fcdf129530b83f00e; paper-result-b31f8dc93eb733d9c3; paper-result-889ce61bcef60c0f8c; paper-result-3b51165690d6d453ee; paper-result-c785e6f619ae381da7; paper-result-7b019b2b8bb9da3d29; paper-result-725d822668ea6d7acd; paper-result-d7a88bfa84a1748144; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 4 SPOT-RNA, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 5 MXfold2, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 6 ContextFold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 7 CONTRAfold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 8 EternaFold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 9 LinearFold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 10 RNAfold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 11 SimFold, column 5: bpRNA-TS0 Recall; Table 1 (Tab1), row 12 RNAstructure, column 5: bpRNA-TS0 Recall; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-archiveii-inf; title: ArchiveII · INF; protocol id: paper-protocol-e8a1aa302bb8df1efd; dataset id: paper-dataset-be2c18a91718ba4592; metric: INF; unit: unitless; direction: higher; result ids: paper-result-0201b139fce549eab3; paper-result-74f904c36aab24aa79; paper-result-4800ab44874448c8b4; paper-result-fcd63eea1286d760ab; paper-result-e6a6c3931d9266f05f; paper-result-9b4f92609217cc1fb6; paper-result-9110b71d1f8e197959; paper-result-ded0df417ed95f9f29; paper-result-e32db43eda24866957; paper-result-45847f62623b463cf2; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 6: ArchiveII INF; Table 1 (Tab1), row 4 SPOT-RNA, column 6: ArchiveII INF; Table 1 (Tab1), row 5 MXfold2, column 6: ArchiveII INF; Table 1 (Tab1), row 6 ContextFold, column 6: ArchiveII INF; Table 1 (Tab1), row 7 CONTRAfold, column 6: ArchiveII INF; Table 1 (Tab1), row 8 EternaFold, column 6: ArchiveII INF; Table 1 (Tab1), row 9 LinearFold, column 6: ArchiveII INF; Table 1 (Tab1), row 10 RNAfold, column 6: ArchiveII INF; Table 1 (Tab1), row 11 SimFold, column 6: ArchiveII INF; Table 1 (Tab1), row 12 RNAstructure, column 6: ArchiveII INF; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-archiveii-f1; title: ArchiveII · F1; protocol id: paper-protocol-e8a1aa302bb8df1efd; dataset id: paper-dataset-be2c18a91718ba4592; metric: F1; unit: unitless; direction: higher; result ids: paper-result-848e79a6dda494f6d2; paper-result-15bf14f0627ace6e51; paper-result-b261f94a241a11866c; paper-result-f6296b4b6029ae0c24; paper-result-4966fd60b18e79a44a; paper-result-3f5f78f9847ba7def0; paper-result-64436b2115ee5881b4; paper-result-145bc6b041b9d5dbe4; paper-result-61f86306875259316e; paper-result-8d618367d005102a41; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 7: ArchiveII F1; Table 1 (Tab1), row 4 SPOT-RNA, column 7: ArchiveII F1; Table 1 (Tab1), row 5 MXfold2, column 7: ArchiveII F1; Table 1 (Tab1), row 6 ContextFold, column 7: ArchiveII F1; Table 1 (Tab1), row 7 CONTRAfold, column 7: ArchiveII F1; Table 1 (Tab1), row 8 EternaFold, column 7: ArchiveII F1; Table 1 (Tab1), row 9 LinearFold, column 7: ArchiveII F1; Table 1 (Tab1), row 10 RNAfold, column 7: ArchiveII F1; Table 1 (Tab1), row 11 SimFold, column 7: ArchiveII F1; Table 1 (Tab1), row 12 RNAstructure, column 7: ArchiveII F1; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-archiveii-precision; title: ArchiveII · Precision; protocol id: paper-protocol-e8a1aa302bb8df1efd; dataset id: paper-dataset-be2c18a91718ba4592; metric: Precision; unit: unitless; direction: higher; result ids: paper-result-cab7a52a6159f4b3f5; paper-result-4e6b9de9fa6581febe; paper-result-4e335c3e3fc60bd7a3; paper-result-0d6f9ea5b80a61c517; paper-result-6fd617ba744c55f823; paper-result-80618ad04474cfdfa4; paper-result-d0f2c61c9603a219fd; paper-result-2bf7f6f35cb278b8b3; paper-result-b4a0a133d3c57bbd78; paper-result-af213ccd00f688f0e8; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 8: ArchiveII Precision; Table 1 (Tab1), row 4 SPOT-RNA, column 8: ArchiveII Precision; Table 1 (Tab1), row 5 MXfold2, column 8: ArchiveII Precision; Table 1 (Tab1), row 6 ContextFold, column 8: ArchiveII Precision; Table 1 (Tab1), row 7 CONTRAfold, column 8: ArchiveII Precision; Table 1 (Tab1), row 8 EternaFold, column 8: ArchiveII Precision; Table 1 (Tab1), row 9 LinearFold, column 8: ArchiveII Precision; Table 1 (Tab1), row 10 RNAfold, column 8: ArchiveII Precision; Table 1 (Tab1), row 11 SimFold, column 8: ArchiveII Precision; Table 1 (Tab1), row 12 RNAstructure, column 8: ArchiveII Precision; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab1-archiveii-recall; title: ArchiveII · Recall; protocol id: paper-protocol-e8a1aa302bb8df1efd; dataset id: paper-dataset-be2c18a91718ba4592; metric: Recall; unit: unitless; direction: higher; result ids: paper-result-23d3090d9e3540349d; paper-result-b79992ef838378f5c1; paper-result-e038536d6f9b001984; paper-result-d82764605cdb73aad8; paper-result-57de4fbe015fbae501; paper-result-a8cf704b852f820e32; paper-result-5f273213a7f48d1b55; paper-result-8164922a733ea800eb; paper-result-ccd93245b92c3d3620; paper-result-cf786103471c989dd3; source ids: bpfold-2025; source locator: Sequence-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold), and non-ML methods (LinearFold, RNAfold, SimFold, and RNAstructure) on bpRNA-TS0 ( n = 1305 RNAs) and ArchiveII ( n = 3966 RNAs) datasets; Table 1 (Tab1), row 3 BPfold, column 9: ArchiveII Recall; Table 1 (Tab1), row 4 SPOT-RNA, column 9: ArchiveII Recall; Table 1 (Tab1), row 5 MXfold2, column 9: ArchiveII Recall; Table 1 (Tab1), row 6 ContextFold, column 9: ArchiveII Recall; Table 1 (Tab1), row 7 CONTRAfold, column 9: ArchiveII Recall; Table 1 (Tab1), row 8 EternaFold, column 9: ArchiveII Recall; Table 1 (Tab1), row 9 LinearFold, column 9: ArchiveII Recall; Table 1 (Tab1), row 10 RNAfold, column 9: ArchiveII Recall; Table 1 (Tab1), row 11 SimFold, column 9: ArchiveII Recall; Table 1 (Tab1), row 12 RNAstructure, column 9: ArchiveII Recall; context: Sequence-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-rfam12-3-14-10-inf; title: Rfam12.3–14.10 · INF; protocol id: paper-protocol-4c3af10c18f615709d; dataset id: paper-dataset-816ebb930725137655; metric: INF; unit: unitless; direction: higher; result ids: paper-result-bc33c532295af7b251; paper-result-d3521184aa90f4092f; paper-result-625df7ddcc5441ab69; paper-result-ae0e3455f6b89e6f70; paper-result-57fc7701f0c9da77d0; paper-result-26e74e5c77ce1570a9; paper-result-ad4f1a6205fecd1b6d; paper-result-d197148a19effcf018; paper-result-68b0000887cc4b5e80; paper-result-853e1b3f8354ef0290; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 4 SPOT-RNA, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 5 MXfold2, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 6 ContextFold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 7 CONTRAfold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 8 EternaFold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 9 LinearFold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 10 RNAfold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 11 SimFold, column 2: Rfam12.3–14.10 INF; Table 2 (Tab2), row 12 RNAstructure, column 2: Rfam12.3–14.10 INF; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-rfam12-3-14-10-f1; title: Rfam12.3–14.10 · F1; protocol id: paper-protocol-4c3af10c18f615709d; dataset id: paper-dataset-816ebb930725137655; metric: F1; unit: unitless; direction: higher; result ids: paper-result-d8fb52a660fba363da; paper-result-3019f81982dc6332cf; paper-result-0ce94efc1ed30c3769; paper-result-a34fce1d750df98700; paper-result-ac79fb8c90b25f7318; paper-result-3db670da1582c96f89; paper-result-f60a9b4fba0bdd5bca; paper-result-5c3a3c43a76fcaeb84; paper-result-7f0004110bb610598b; paper-result-31c527dabf4e1f6958; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 4 SPOT-RNA, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 5 MXfold2, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 6 ContextFold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 7 CONTRAfold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 8 EternaFold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 9 LinearFold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 10 RNAfold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 11 SimFold, column 3: Rfam12.3–14.10 F1; Table 2 (Tab2), row 12 RNAstructure, column 3: Rfam12.3–14.10 F1; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-rfam12-3-14-10-precision; title: Rfam12.3–14.10 · Precision; protocol id: paper-protocol-4c3af10c18f615709d; dataset id: paper-dataset-816ebb930725137655; metric: Precision; unit: unitless; direction: higher; result ids: paper-result-4c05190191ba52c1d1; paper-result-22774caf06553a0120; paper-result-03b61894f52b9f6a21; paper-result-39a55efc7440a14e65; paper-result-c859a2ad30d5982c35; paper-result-d49a44e7ee0390f25d; paper-result-6e6c7831c443afa6ed; paper-result-e59816ecb58e31617e; paper-result-b6a697b1154c255641; paper-result-f3fb35d7219cc3367e; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 4 SPOT-RNA, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 5 MXfold2, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 6 ContextFold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 7 CONTRAfold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 8 EternaFold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 9 LinearFold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 10 RNAfold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 11 SimFold, column 4: Rfam12.3–14.10 Precision; Table 2 (Tab2), row 12 RNAstructure, column 4: Rfam12.3–14.10 Precision; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-rfam12-3-14-10-recall; title: Rfam12.3–14.10 · Recall; protocol id: paper-protocol-4c3af10c18f615709d; dataset id: paper-dataset-816ebb930725137655; metric: Recall; unit: unitless; direction: higher; result ids: paper-result-399712a182e4786639; paper-result-c824c3c83e5e451637; paper-result-a1f081e3b661ff4c25; paper-result-99b4cc2229fcdd0626; paper-result-0c3b54bf620bf11af3; paper-result-b94255fbaa7ff64d55; paper-result-453956286dc654a2fc; paper-result-30d6b97036164e9290; paper-result-c7d3e722ba9b016c03; paper-result-975ccfae7d5107a44a; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 4 SPOT-RNA, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 5 MXfold2, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 6 ContextFold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 7 CONTRAfold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 8 EternaFold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 9 LinearFold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 10 RNAfold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 11 SimFold, column 5: Rfam12.3–14.10 Recall; Table 2 (Tab2), row 12 RNAstructure, column 5: Rfam12.3–14.10 Recall; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-pdb-inf; title: PDB · INF; protocol id: paper-protocol-b8acf180ccd3e67923; dataset id: reported-dataset-8317793f18b026; metric: INF; unit: unitless; direction: higher; result ids: paper-result-97b54396c089187bac; paper-result-5465552e8e451ecd56; paper-result-e465e8b97a72acc635; paper-result-d1041275466a1cb1ae; paper-result-3d0a9108d41aebb9db; paper-result-3ddc4380f7d4e8cc34; paper-result-738dbc58785cabadb7; paper-result-c2cbc18894bd07f1a6; paper-result-155bc3cd1981f1e9fd; paper-result-cfbfd2a82ace58ed84; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 6: PDB INF; Table 2 (Tab2), row 4 SPOT-RNA, column 6: PDB INF; Table 2 (Tab2), row 5 MXfold2, column 6: PDB INF; Table 2 (Tab2), row 6 ContextFold, column 6: PDB INF; Table 2 (Tab2), row 7 CONTRAfold, column 6: PDB INF; Table 2 (Tab2), row 8 EternaFold, column 6: PDB INF; Table 2 (Tab2), row 9 LinearFold, column 6: PDB INF; Table 2 (Tab2), row 10 RNAfold, column 6: PDB INF; Table 2 (Tab2), row 11 SimFold, column 6: PDB INF; Table 2 (Tab2), row 12 RNAstructure, column 6: PDB INF; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-pdb-f1; title: PDB · F1; protocol id: paper-protocol-b8acf180ccd3e67923; dataset id: reported-dataset-8317793f18b026; metric: F1; unit: unitless; direction: higher; result ids: lit-011; paper-result-57c0ff7b77a652ab1c; paper-result-3d2877af386639e0ad; paper-result-4ae34e1cc310c60554; paper-result-b6a3c4b03b40732ef2; paper-result-96d2c93b67543eaa47; paper-result-9440f5e0b35b26980b; lit-012; paper-result-fda0f45ba9232f7233; paper-result-c35bfeb0f7fc67fda6; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 7: PDB F1; Table 2 (Tab2), row 4 SPOT-RNA, column 7: PDB F1; Table 2 (Tab2), row 5 MXfold2, column 7: PDB F1; Table 2 (Tab2), row 6 ContextFold, column 7: PDB F1; Table 2 (Tab2), row 7 CONTRAfold, column 7: PDB F1; Table 2 (Tab2), row 8 EternaFold, column 7: PDB F1; Table 2 (Tab2), row 9 LinearFold, column 7: PDB F1; Table 2 (Tab2), row 10 RNAfold, column 7: PDB F1; Table 2 (Tab2), row 11 SimFold, column 7: PDB F1; Table 2 (Tab2), row 12 RNAstructure, column 7: PDB F1; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-pdb-precision; title: PDB · Precision; protocol id: paper-protocol-b8acf180ccd3e67923; dataset id: reported-dataset-8317793f18b026; metric: Precision; unit: unitless; direction: higher; result ids: paper-result-8319fb26cf3f85d194; paper-result-eac4aa84532a86e39b; paper-result-b6a55e179b51c6398e; paper-result-5ae72ffe6dc5e67fcd; paper-result-d4877662e8f67af294; paper-result-49bafa3b1c623a5f8d; paper-result-c96dd6eb50a381f576; paper-result-2d080d038082cd24e2; paper-result-ddafb4472501414293; paper-result-c9d2c92f2b018f5062; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 8: PDB Precision; Table 2 (Tab2), row 4 SPOT-RNA, column 8: PDB Precision; Table 2 (Tab2), row 5 MXfold2, column 8: PDB Precision; Table 2 (Tab2), row 6 ContextFold, column 8: PDB Precision; Table 2 (Tab2), row 7 CONTRAfold, column 8: PDB Precision; Table 2 (Tab2), row 8 EternaFold, column 8: PDB Precision; Table 2 (Tab2), row 9 LinearFold, column 8: PDB Precision; Table 2 (Tab2), row 10 RNAfold, column 8: PDB Precision; Table 2 (Tab2), row 11 SimFold, column 8: PDB Precision; Table 2 (Tab2), row 12 RNAstructure, column 8: PDB Precision; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: bpfold-2025-tab2-pdb-recall; title: PDB · Recall; protocol id: paper-protocol-b8acf180ccd3e67923; dataset id: reported-dataset-8317793f18b026; metric: Recall; unit: unitless; direction: higher; result ids: paper-result-5afb76a439dd9677b9; paper-result-3ad0144bc78eae6759; paper-result-01efc77bc3fe0c8593; paper-result-5fe58d35cc68e79078; paper-result-11126a2ff54dbf3afb; paper-result-a45e9b5b53f964bd93; paper-result-8828c8e20f56d3d354; paper-result-6f6f94f1d2e661baac; paper-result-a7667b03bef3a87b94; paper-result-c0cf50ee8c90573ad2; source ids: bpfold-2025; source locator: Family-wise evaluation of three DL methods (BPfold, SPOT-RNA, and MXfold2), three shallow learning methods (ContextFold, CONTRAfold, and EternaFold) and non-learning methods (LinearFold, RNAfold, SimFold, and RNAstructure) on Rfam12.3–14.10 ( n = 10,791 RNAs) and PDB ( n = 116 RNAs) datasets; Table 2 (Tab2), row 3 BPfold, column 9: PDB Recall; Table 2 (Tab2), row 4 SPOT-RNA, column 9: PDB Recall; Table 2 (Tab2), row 5 MXfold2, column 9: PDB Recall; Table 2 (Tab2), row 6 ContextFold, column 9: PDB Recall; Table 2 (Tab2), row 7 CONTRAfold, column 9: PDB Recall; Table 2 (Tab2), row 8 EternaFold, column 9: PDB Recall; Table 2 (Tab2), row 9 LinearFold, column 9: PDB Recall; Table 2 (Tab2), row 10 RNAfold, column 9: PDB Recall; Table 2 (Tab2), row 11 SimFold, column 9: PDB Recall; Table 2 (Tab2), row 12 RNAstructure, column 9: PDB Recall; context: Family-wise RNA secondary-structure evaluation; macro-average canonical base-pair metrics.; caveats: Within each dataset and metric only. Distinct training data, thermodynamic priors and learned methods are retained; dataset size is not a base-pair denominator. These are BPfold-author evaluations of external methods, not Rewire reproductions. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17
benchmark research: review date: 2026-09-17; status: complete_tables_extracted; primary sources: evidence-expansion-bpfold-2025-976218bd; inspected locators: Table 1; XML table Tab1; Table 2; XML table Tab2; searched queries: Deep generalizable prediction of RNA secondary structure via base pair motif energy 10.1038/s41467-025-60048-1; gaps: 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: bpfold-2025; source locator: Introduction; Methods: Datasets and evaluation; cached text lines 9, 44, 74–77; matching task comparison table/ablation captions; 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: PDB RNA set
evaluates task: bpRNA-TS0 (RNA secondary structure)
evaluates task: Rfam12.3–14.10 (RNA secondary structure)
evaluates task: PDB (RNA secondary structure)
evaluates task: ArchiveII (RNA secondary structure)

Suggest a correction