AUPRC (linear head on frozen pair embeddings plus their cosine similarity, genus-level split, mean over held-out test genomes) of ProkBERT (frozen embeddings with a trained linear head) on BacBench protein-protein interaction dataset from STRING DB
34.38 ± 9.31 AUPRC (linear head on frozen pair embeddings plus their cosine similarity, genus-level split, mean over held-out test genomes)
Methods
- Tested configuration
- ProkBERT (frozen embeddings with a trained linear head)
- Protocol
- BacBench protein-protein interaction, linear probe AUPRC
- Dataset
- BacBench protein-protein interaction dataset from STRING DB
- Procedure
- Zero-shot columns score each protein pair by the cosine similarity of the two embeddings. The finetuned columns train a single linear layer, preceded by layer normalisation and dropout of 0.2, on the average of the two frozen protein embeddings together with their cosine similarity, so the input is of dimension D plus one; binary cross-entropy loss, Adam with weight decay 0.01, at most 20 epochs with early stopping patience 5 on AUROC, and a learning-rate sweep per model. Metrics are computed per genome on the test split.
- Evaluation
- ProkBERT on BacBench protein-protein interaction, linear probe AUPRC
- Coverage
- Not reported scored / Not reported eligible
- Uncertainty
- SD 9.31
- Evidence
- Author-reported evaluation · source checkedBacformer preprint, bioRxiv 2025.07.20.665723 version 3 · Table S13, row 'ProkBERT', column 'Finetuned AUPRC'
A source-checked result verifies the numerical transcription, not every model or protocol detail. Evaluation metadata: source checked. Source checked does not mean independently reproduced.
Reproduction
- Split
- Phylogeny-aware split by genus, 60/20/20 train, validation and test.
- Adaptation
- Linear head trained on frozen embeddings
- Scoring implementation
- linear head on frozen pair embeddings plus their cosine similarity, genus-level split, mean over held-out test genomes
No execution recipe has been verified for this exact configuration and evaluation. A benchmark's general instructions may use different inputs, splits or model settings.
Reproducing this published result requires matching its model configuration, data, split and scorer. Source checking or a successful smoke test does not establish score reproduction.
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.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Reported result 34.38 ± 9.31 Individual claims | Bacformer preprint, bioRxiv 2025.07.20.665723 version 3 Table S13, row 'ProkBERT', column 'Finetuned AUPRC' Version: Version 3, posted 17 August 2026 according to the page banner; the bioRxiv API lists version 3 with date 2026-08-19 | source checked ["independent-cell-check","deterministic-table-parse","pdf-text-parse","source-hash-verification","ai-assisted-source-review"] · 2026-11-15T23:55:00Z author reported Audit detailsProtein-protein interaction, Table S13. The zero-shot and finetuned column group was checked geometrically against the printed group header, not inferred from column order. The unit concept was changed from percent to unitless in this review, with unit_detail 'printed x100'. data/vocab/unit.ttl says so in the scope note of unitless, and percent would assert that an AUROC, an R squared or a Pearson correlation is a proportion of items. The printed value, the numeric value and the spread are unchanged. Field: Source artifact SHA-256: Hash scope: The hash is of one download. bioRxiv regenerates this PDF on request, so the container bytes differ between downloads: a second download minutes later gave 08fcc83af256845d942345fc6b63c64911878733acd028edab5e9307d63319db. The pdftotext -layout text layer of both downloads is byte-identical (SHA-256 86636e89b167d7304b3a83ee7f159cad0a9bdbe97019588ce836087b8fe3b0b2), and that text layer, archived gzipped in artifacts/ with SHA-256 0afdefd78a410d036d1c4f17e33bdf481e8d76241ee02f4c84c7c4be91ee72aa, is what every extracted value was read from. Extraction artifact SHA-256: |
Sources and history
Release 2026-11-21-c8bea25e5abb · Record review: source checked
1 source record and release history
- Bacformer preprint, bioRxiv 2025.07.20.665723 version 3 · Original source · Version 3, posted 17 August 2026 according to the page banner; the bioRxiv API lists version 3 with date 2026-08-19
Technical metadata and extraction receipts
Stable ID: bacformer-20261115-result-ppi-prokbert-probe-auprc
- metric
- auprc
- metric direction
- higher
- metric qualifier
- linear head on frozen pair embeddings plus their cosine similarity, genus-level split, mean over held-out test genomes
- numeric value
- 34.38
- printed value
- 34.38 ± 9.31
- review
- artifact sha256: 86636e89b167d7304b3a83ee7f159cad0a9bdbe97019588ce836087b8fe3b0b2; method: independent-cell-check; deterministic-table-parse; pdf-text-parse; source-hash-verification; ai-assisted-source-review; method note: Independent review pass. Every cell of the six benchmark tables was re-read by review/recheck_cells.py, written for this pass, which uses neither of the extractor's scripts and does not read the pdftotext -layout text layer they read. It takes the word bounding boxes pdftotext -bbox-layout emits for a freshly downloaded version 3 PDF, groups words into printed lines by their y coordinate, derives each table's column boundaries from the x extent of that table's own printed header words, and assigns every number to a column by where it sits on the page, so a value printed under one header cannot be read as a value of another. It asserts the printed row order, that every column of every row is filled, that each cell holds exactly a mean and a spread, that each cell prints its own plus-minus sign, and for Tables S13 and S18 that the group header sits geometrically over the two columns this review names after it. All 186 cells agreed on the printed value, the mean, the spread, the spread's printed string and its source column, with no disagreement, and the six tables print no cell the batch leaves unrecorded. Three negative controls (two cells of one row exchanged, one cell taking its neighbour's value, every Table S18 locator moved one column right) were all rejected, and the unmutated batch accepted; the log is review/negative-controls.txt. npm run batch -- check-values reproduced the extractor's figures exactly, 36 in_cell and 150 in_row with nothing else, against a text layer regenerated from a fresh download. A manual sample of 47 cells was then read by eye against the printed table bodies: 24 drawn with random.sample over all 186 at seed 20261115, all 18 cells of Table S12, and the eight Evo and Evo-2 cells of Tables S1 and S14, where the printed row order differs from the other tables. All matched, column assignment included.; note: Protein-protein interaction, Table S13. The zero-shot and finetuned column group was checked geometrically against the printed group header, not inferred from column order. The unit concept was changed from percent to unitless in this review, with unit_detail 'printed x100'. data/vocab/unit.ttl says so in the scope note of unitless, and percent would assert that an AUROC, an R squared or a Pearson correlation is a proportion of items. The printed value, the numeric value and the spread are unchanged.; retrieval url: https://www.biorxiv.org/content/10.1101/2025.07.20.665723v3.full.pdf; reviewed at: 2026-11-15T23:55:00Z; reviewer: claude; reviewer note: Claude (Opus 5) review agent, independent of the agent that extracted this batch; no human review claimed. Transcription was checked against the pinned sources. Nothing was reproduced and no value was recomputed.
- source locator
- Table S13, row 'ProkBERT', column 'Finetuned AUPRC'
- source table
- S13
- uncertainty
- printed: 34.38 ± 9.31; scope: across held-out test genomes; source column: Finetuned AUPRC; type: standard_deviation; unit: unitless; value: 9.31
- unit
- unitless
- unit detail
- printed x100