NCBench: providing an open, reproducible, transparent, adaptable, and continuous benchmark approach for DNA-sequencing-based variant calling
Bibliographic record
Abstract
We present the results of the human genomic small variant calling benchmarking initiative of the German Research Foundation (DFG) funded Next Generation Sequencing Competence Network (NGS-CN) and the German Human Genome-Phenome Archive (GHGA). In this effort, we developed NCBench, a continuous benchmarking platform for the evaluation of small genomic variant callsets in terms of recall, precision, and false positive/negative error patterns. NCBench is implemented as a continuously re-evaluated open-source repository. We show that it is possible to entirely rely on public free infrastructure (Github, Github Actions, Zenodo) in combination with established open-source tools. NCBench is agnostic of the used dataset and can evaluate an arbitrary number of given callsets, while reporting the results in a visual and interactive way. We used NCBench to evaluate over 40 callsets generated by various variant calling pipelines available in the participating groups that were run on three exome datasets from different enrichment kits and at different coverages. While all pipelines achieve high overall quality, subtle systematic differences between callers and datasets exist and are made apparent by NCBench.These insights are useful to improve existing pipelines and develop new workflows. NCBench is meant to be open for the contribution of any given callset. Most importantly, for authors, it will enable the omission of repeated re-implementation of paper-specific variant calling benchmarks for the publication of new tools or pipelines, while readers will benefit from being able to (continuously) observe the performance of tools and pipelines at the time of reading instead of at the time of writing.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.027 | 0.069 |
| Meta-epidemiology (narrow) | 0.004 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.007 | 0.006 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.007 | 0.004 |
| Open science | 0.010 | 0.007 |
| Research integrity | 0.003 | 0.004 |
| Insufficient payload (model declined to judge) | 0.007 | 0.009 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".