GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
Splice site prediction, dataset Reconstruct. Scored with MCC on GUE Splice site prediction, Reconstruct. Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12.
Overview
Splice site prediction, dataset Reconstruct. Scored with MCC on GUE Splice site prediction, Reconstruct. Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12.
Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.
Evaluation design
Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.
Benchmarks
These source-backed links do not make different protocols or scores interchangeable.
Recorded evaluations
Each evaluation records what was tested and under which conditions.
- DNABERT-2 (further pre-trained on GUE) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- DNABERT-2 on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- DNABERT (3-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- DNABERT (4-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- DNABERT (5-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- DNABERT (6-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- NT-2500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- NT-2500M-multi on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- NT-500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- NT-500M-human on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
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.
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. The pooled view gathers every source table that reports the same metric and names what it does not hold constant.
GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
mcc (percent) · Higher values are better for this metric.
Every method GUE reports on Splice site prediction, dataset Reconstruct, scored with MCC on GUE Splice site prediction, Reconstruct.
Evaluation protocol · GUE Splice site prediction, Reconstruct (GUE split)
- DNABERT (3-mer) · Configuration · Author-reported evaluation84.14
- DNABERT (4-mer) · Configuration · Author-reported evaluation84.05
- DNABERT (5-mer) · Configuration · Author-reported evaluation84.02
- DNABERT (6-mer) · Configuration · Author-reported evaluation84.07
- NT-500M-human · Configuration · Author-reported evaluation79.71
- NT-500M-1000g · Configuration · Author-reported evaluation80.97
- NT-2500M-1000g · Configuration · Author-reported evaluation85.78
- NT-2500M-multi · Configuration · Author-reported evaluation89.35
- DNABERT-2 · Configuration · Author-reported evaluation84.99
- DNABERT-2 (further pre-trained on GUE) · Configuration · Author-reported evaluation85.93
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.
DNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 12, row(Splice site prediction)Values, uncertainty and evidence
| Tested entity | Printed value | Uncertainty | Evidence |
|---|---|---|---|
| DNABERT (3-mer) · Configuration | 84.14 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (3-mer)), column(Splice site prediction Reconstruct) |
| DNABERT (4-mer) · Configuration | 84.05 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (4-mer)), column(Splice site prediction Reconstruct) |
| DNABERT (5-mer) · Configuration | 84.02 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (5-mer)), column(Splice site prediction Reconstruct) |
| DNABERT (6-mer) · Configuration | 84.07 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (6-mer)), column(Splice site prediction Reconstruct) |
| NT-500M-human · Configuration | 79.71 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-500M-human), column(Splice site prediction Reconstruct) |
| NT-500M-1000g · Configuration | 80.97 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-500M-1000g), column(Splice site prediction Reconstruct) |
| NT-2500M-1000g · Configuration | 85.78 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-2500M-1000g), column(Splice site prediction Reconstruct) |
| NT-2500M-multi · Configuration | 89.35 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-2500M-multi), column(Splice site prediction Reconstruct) |
| DNABERT-2 · Configuration | 84.99 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT-2), column(Splice site prediction Reconstruct) |
| DNABERT-2 (further pre-trained on GUE) · Configuration | 85.93 percent | Not reported | Author-reported evaluation · source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT-2♦), column(Splice site prediction Reconstruct) |
Scope and limitations
- Author-reported numbers, source checked but not independently reproduced.
- Scores are MCC, except Covid variant classification which is F1, both on a 0 to 100 scale.
- The diamond entry is DNABERT-2 with further pre-training on the GUE training sets, so it is not directly comparable to the others.
Source transcription and grouping reviewed by automated source review on 2026-09-18. These experiments were not independently reproduced by rewire.
Tested entities and results
Release 2026-09-17-134cd1815de8 · 10 evaluations · 10 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.
| Metric and finding | Coverage and uncertainty | Evidence |
|---|---|---|
| DNABERT-2 (further pre-trained on GUE) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT-2 (further pre-trained on GUE)Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 85.93% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT-2♦), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| DNABERT-2 on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT-2Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 84.99% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT-2), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| DNABERT (3-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT (3-mer)Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 84.14% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (3-mer)), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| DNABERT (4-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT (4-mer)Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 84.05% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (4-mer)), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| DNABERT (5-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT (5-mer)Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 84.02% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (5-mer)), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| DNABERT (6-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: DNABERT (6-mer)Task: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 84.07% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(DNABERT (6-mer)), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| NT-2500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: NT-2500M-1000gTask: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 85.78% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-2500M-1000g), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| NT-2500M-multi on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: NT-2500M-multiTask: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 89.35% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-2500M-multi), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| NT-500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: NT-500M-1000gTask: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 80.97% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-500M-1000g), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
| NT-500M-human on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct Configuration: NT-500M-humanTask: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset ReconstructDataset subset: GUE Splice site prediction, Reconstruct (GUE split) Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12. Author-reported evaluation · Evaluation metadata: source checked | ||
| 79.71% mcc Unit: percent · Direction: higher | Uncertainty: Not reported Scored: Not reported · Eligible: Not reported | source checkedDNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Table 6, row(NT-500M-human), column(Splice site prediction Reconstruct) Source checking is not independent reproduction. |
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Relationship: part of discovery-benchmark-gue Individual claims | DNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes Table 12, row(Splice site prediction) Version: Primary full-text snapshot retrieved 2026-09-17; exact bytes pinned by SHA-256 | source checked automated source review · 2026-09-18 Audit detailsPrimary-source transcription with no human sign-off and no independent reproduction. Field: Claim: gue-association-splice-site-prediction-reconstruct Source artifact SHA-256: Hash scope: Exact retrieved primary paper artifact bytes. |
Sources and history
Release 2026-09-17-134cd1815de8 · Record review: source checked
1 source records and release history
- DNABERT-2: Efficient Foundation Model and Benchmark for Multi-Species Genomes · Original source · Primary full-text snapshot retrieved 2026-09-17; exact bytes pinned by SHA-256
Technical metadata and extraction receipts
Stable ID: gue-task-splice-site-prediction-reconstruct
- areas
- dna-genomes
- tasks
- Splice site prediction, dataset Reconstruct
- metric
- MCC
- metric direction
- higher
- dataset
- GUE Splice site prediction, Reconstruct
- protocol
- Fine-tuned on the GUE training split, scored on its test split. Split sizes are in Table 12.
- source locator
- Table 12, row(Splice site prediction)
- comparison panels
- id: gue-panel-splice-site-prediction-reconstruct; title: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct; protocol id: gue-task-splice-site-prediction-reconstruct; dataset id: gue-dataset-gue-splice-site-prediction-reconstruct; metric: mcc; unit: percent; direction: higher; result ids: gue-result-dnabert-3-mer-splice-site-prediction-reconstruct-mcc; gue-result-dnabert-4-mer-splice-site-prediction-reconstruct-mcc; gue-result-dnabert-5-mer-splice-site-prediction-reconstruct-mcc; gue-result-dnabert-6-mer-splice-site-prediction-reconstruct-mcc; gue-result-nt-500m-human-splice-site-prediction-reconstruct-mcc; gue-result-nt-500m-1000g-splice-site-prediction-reconstruct-mcc; gue-result-nt-2500m-1000g-splice-site-prediction-reconstruct-mcc; gue-result-nt-2500m-multi-splice-site-prediction-reconstruct-mcc; gue-result-dnabert-2-splice-site-prediction-reconstruct-mcc; gue-result-dnabert-2-further-pre-trained-on-gue-splice-site-prediction-reconstruct-mcc; source ids: evidence-expansion-gue-49300ace; source locator: Table 12, row(Splice site prediction); context: Every method GUE reports on Splice site prediction, dataset Reconstruct, scored with MCC on GUE Splice site prediction, Reconstruct.; caveats: Author-reported numbers, source checked but not independently reproduced.; Scores are MCC, except Covid variant classification which is F1, both on a 0 to 100 scale.; The diamond entry is DNABERT-2 with further pre-training on the GUE training sets, so it is not directly comparable to the others.; review: method: automated_source_review; date: 2026-09-18
Related records
- part of: GUE
- subject: GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: part of discovery-benchmark-gue
- benchmark: DNABERT-2 (further pre-trained on GUE) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: DNABERT-2 on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: DNABERT (3-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: DNABERT (4-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: DNABERT (5-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: DNABERT (6-mer) on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: NT-2500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: NT-2500M-multi on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: NT-500M-1000g on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct
- benchmark: NT-500M-human on GUE SPLICE-SITE-PREDICTION-RECONSTRUCT: Splice site prediction, dataset Reconstruct