A high-dimensional benchmark of objective functions and biological resolutions for personalized metabolic phenotyping
Mehmet Ali Erdoğan, Ali Cakmak · 2026
Read A high-dimensional benchmark of objective functions and biological resolutions for personalized metabolic phenotyping by Mehmet Ali Erdoğan, Ali Cakmak (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13625657.txt.
Description
Mehmet Ali Erdoğan and Ali Cakmak benchmark how to turn personal metabolic models into diagnostic signals in a 2026 paper. Genome-scale models simulate a person's internal chemistry, but someone must choose what the model optimizes, at what biological resolution, and which features to feed a classifier. The team treated the learning algorithm as a controlled backdrop and varied the rest.
Over fifty-seven thousand configurations spanned six diseases, from cancer to neurodegeneration. Pathway-level minimum flux, a bottleneck-style measure, carried the strongest signal, beating aggregate pathway activity. Accuracy barely moved with feature count under leakage-free testing, so aggressive feature pruning was needless. Most classifiers were robust at full density; only one sagged. Adding objective-function values back as features recovered context lost in aggregation. The benchmark gives modelers an evidence-based default recipe.
Catalog record
| CREATOR | Mehmet Ali Erdoğan, Ali Cakmak |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |