Large-scale Manual Curation and Harmonization of Metadata from Metagenomic and Cancer Genomic Repositories: Challenges and Solutions
Bibliographic record
Abstract
Public omics repositories contain vast amounts of valuable data, but their metadata suffers from extreme heterogeneity, unstandardized terminologies, and quality issues that severely limit data reusability and cross-study integration. While prospective metadata standards exist, the majority of published omics data remain in non-standardized formats requiring retrospective harmonization. We performed comprehensive manual curation and harmonization of metadata, such as participant characteristics and study conditions, from 212 027 omics samples across 468 studies in two repositories: curatedMetagenomicData (93 studies, 22 588 samples) and cBioPortal (375 studies, 189 438 samples). Through systematic ontology mapping, we consolidated redundant, dispersed information into far fewer harmonized columns, reduced unique values, and increased the completeness of major attributes. This curation process revealed common metadata quality issues, including typos, inconsistent terminologies, misplaced values, conflicting annotations, and inappropriately merged information across attributes. We document the challenges, decisions, and solutions during this large-scale metadata harmonization. The harmonized metadata, accessible through the OmicsMLRepoR Bioconductor package, enables repository-wide queries and cross-study analyses previously challenging with heterogeneous metadata. Our experience provides practical guidance for similar curation efforts and demonstrates the value of investing in retrospective metadata improvement for existing public omics resources.
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 | 0.187 | 0.265 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.018 | 0.022 |
| Science and technology studies | 0.008 | 0.005 |
| Scholarly communication | 0.014 | 0.012 |
| Open science | 0.008 | 0.021 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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".