Machine learning on pyrolysis GC-MS data separates biotic from abiotic samples with about 90 percent accuracy
Source check: Needs source check. Not independent scientific confirmation.
Finding: Cleaves et al. trained machine-learning models on pyrolysis-GC-MS data from living cells, fossil organic material, carbon-rich meteorites and synthetic organics (134 samples per secondary sources) to classify biogenicity of contemporary and ancient samples at about 90 percent accuracy without identifying individual compounds. Accuracy comes from held-out data within the same sample set; no blind or flight validation and no detection of extraterrestrial life is reported. Abstract-level review.
Technology: Supervised machine-learning classifiers on chromatographic and mass-spectral peak patterns
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Models learn relational patterns in peak data to classify biogenicity.
- Test and result
- Train and test subsets of the sample set gave about 90 percent accuracy (per abstract).
- Limitation / still unknown
- Independent samples, model identity and sample count were not verified from the paper itself; generalisation to alien biochemistry is a stated aim, not a demonstrated result.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://europepmc.org/article/MED/37748080
Needs source check · editorial details
The source has not been fully checked or contains conflicting information. 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: 17 Nov 2026. Review scope: abstract and bibliographic metadata (Europe PMC) only; Methods and full text not assessed.
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.