SpliceAI2 training split and benchmark overlap
Training split of SpliceAI2
Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training.
Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training.
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
12 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Claimed field training_split Context-only references | A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: SHA-256 of one retrieval of the bioRxiv PDF. bioRxiv stamps the PDF when serving it (XMP ModifyDate, MetadataDate, DocumentID, InstanceID, Info ModDate and the second trailer ID), so later retrievals differ. A reviewer retrieval on 2026-10-13 of the same length (2,679,664 bytes) had SHA-256 407f5884427dda0226e0ce1f022e4479d3ff2b243da18bcc691eb345b0e04a6a and ModDate D:20261013063115-07'00'; its bytes differ only in those fields and its pdftotext text layer is identical to the pinned copy. |
| Claimed field training_split Context-only references | Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv version 1 supplement DC1 (757673_file02.tsv) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Source location Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Context-only references | A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: SHA-256 of one retrieval of the bioRxiv PDF. bioRxiv stamps the PDF when serving it (XMP ModifyDate, MetadataDate, DocumentID, InstanceID, Info ModDate and the second trailer ID), so later retrievals differ. A reviewer retrieval on 2026-10-13 of the same length (2,679,664 bytes) had SHA-256 407f5884427dda0226e0ce1f022e4479d3ff2b243da18bcc691eb345b0e04a6a and ModDate D:20261013063115-07'00'; its bytes differ only in those fields and its pdftotext text layer is identical to the pinned copy. |
| Source location Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Context-only references | Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv version 1 supplement DC1 (757673_file02.tsv) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Claimed value Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training. Context-only references | A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: SHA-256 of one retrieval of the bioRxiv PDF. bioRxiv stamps the PDF when serving it (XMP ModifyDate, MetadataDate, DocumentID, InstanceID, Info ModDate and the second trailer ID), so later retrievals differ. A reviewer retrieval on 2026-10-13 of the same length (2,679,664 bytes) had SHA-256 407f5884427dda0226e0ce1f022e4479d3ff2b243da18bcc691eb345b0e04a6a and ModDate D:20261013063115-07'00'; its bytes differ only in those fields and its pdftotext text layer is identical to the pinned copy. |
| Claimed value Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training. Context-only references | Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv version 1 supplement DC1 (757673_file02.tsv) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Description Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training. Context-only references | A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: SHA-256 of one retrieval of the bioRxiv PDF. bioRxiv stamps the PDF when serving it (XMP ModifyDate, MetadataDate, DocumentID, InstanceID, Info ModDate and the second trailer ID), so later retrievals differ. A reviewer retrieval on 2026-10-13 of the same length (2,679,664 bytes) had SHA-256 407f5884427dda0226e0ce1f022e4479d3ff2b243da18bcc691eb345b0e04a6a and ModDate D:20261013063115-07'00'; its bytes differ only in those fields and its pdftotext text layer is identical to the pinned copy. |
| Description Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training. Context-only references | Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv version 1 supplement DC1 (757673_file02.tsv) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Relationship: subject rna-splicing-20261013-spliceai2-model-spliceai2 Context-only references | A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: SHA-256 of one retrieval of the bioRxiv PDF. bioRxiv stamps the PDF when serving it (XMP ModifyDate, MetadataDate, DocumentID, InstanceID, Info ModDate and the second trailer ID), so later retrievals differ. A reviewer retrieval on 2026-10-13 of the same length (2,679,664 bytes) had SHA-256 407f5884427dda0226e0ce1f022e4479d3ff2b243da18bcc691eb345b0e04a6a and ModDate D:20261013063115-07'00'; its bytes differ only in those fields and its pdftotext text layer is identical to the pinned copy. |
| Relationship: subject rna-splicing-20261013-spliceai2-model-spliceai2 Context-only references | Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens' Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: bioRxiv version 1 supplement DC1 (757673_file02.tsv) | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-10-22-1b7cf027e4f6 · Record review: source checked
2 source records and release history
- A unified framework for quantitative splicing and transcript prediction (Jaganathan et al., bioRxiv 2026) · Original source · bioRxiv preprint version 1, posted 2026-10-11; full-text PDF (57 pages)
- Supplementary Table 1: training RNA-seq samples and junctions per species (Jaganathan et al., bioRxiv 2026) · Original source · bioRxiv version 1 supplement DC1 (757673_file02.tsv)
Technical metadata and extraction receipts
Stable ID: rna-splicing-20261013-spliceai2-claim-training-split
- field
- training_split
- value
- Genes on human chromosomes 21 and 22 were held out for validation and all other genes were used for training. The variant benchmarks (GTEx and GEL GT/AG, GTEx sQTL, OpenSplice, UK Biobank) are not restricted to held-out chromosomes, so most benchmark variants lie in genes seen in training; only the top-transcript benchmark is restricted to chromosomes 21 and 22. The paper does not say whether junctions observed in GTEx benchmark samples were excluded from training.
- source locator
- Methods 'SpliceAI2 model' (Training, last paragraph), 'Top transcript benchmark' and 'Variant effect benchmarks'; Supplementary Table 1 row 'Homo_sapiens'
- review
- method: pdf-text-parse; independent-cell-check; reviewer: claude; reviewer note: Claude (Opus 5.5) research agent, independent reviewer; did not extract this batch; no human review claimed; date: 2026-10-13; artifact sha256: f8f3db2c3943d4827b670bc2e962e0cec8fbe96e0b1acf9d4a9b9f0c7a78c06b; retrieval url: https://www.biorxiv.org/content/10.64898/2026.10.08.757673v1.full.pdf; method note: Sentence located in the reviewer's own pdftotext output of the pinned PDF; printed value and its context re-read.; note: Checked against the pinned source by the independent reviewer; verdict and correction list in docs/reviews/use-cases/spliceai2-2026-10-13.md and this batch's review.json. Developer-reported value, not reproduced. Extraction method: Split sentence asserted verbatim; the overlap statement is the extractor's reading of the Methods and needs reviewer confirmation.
Related records
- subject: SpliceAI2