Are we there yet? A machine learning architecture to predict organotropic metastases
Source check: Source checked. Not independent scientific confirmation.
Finding: A tree-based machine-learning architecture classified metastatic sites from primary-tumor RNA profiles and identified overlapping enriched processes across several metastatic destinations. The analysis uses retrospective TCGA data and is limited to eligible cancer types and annotated cases; it does not establish prospective prediction.
Technology: Statistical feature selection and tree-based classifiers for metastatic-site prediction from RNA profiles
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Machine-learning classifiers use primary-tumor transcript features to predict metastatic sites; selected features support pathway-enrichment analysis.
- Test and result
- Across 35 cancer-site pairs, the study reported mean precision, recall, F1 and accuracy of 0.82 and overlapping enriched processes for four metastatic sites.
- Limitation / still unknown
- A 30% holdout from TCGA is retrospective internal validation; findings cover 16 of 33 studied cancer types and depend on annotated profiles.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12920-021-01122-7
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary full-text Methods and Results; publication date and catalog comparisons checked..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.