A Bayesian framework for multivariate differential analysis
Marie Chion, Arthur Leroy, Ilya Ioshikhes, Lun Hu, Ilya Ioshikhes, Ilya Ioshikhes · 2026
Read A Bayesian framework for multivariate differential analysis by Marie Chion, Arthur Leroy, Ilya Ioshikhes, Lun Hu, Ilya Ioshikhes, Ilya Ioshikhes (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-13561429.txt.
Description
Marie Chion and Arthur Leroy propose a fully Bayesian approach to differential analysis, the routine hunt for changes between conditions. The workhorse tool, limma, is fast and cheap, but standard testing tends to ignore missing values, correlations between elements, and uncertainty. The authors build a layered model with paired closed forms for each element's mean and variance.
Because the forms are conjugate, the math stays in closed form and costs about as much as ordinary testing. The method samples the difference between conditions, giving calibrated uncertainty. An extension handles many elements at once, respecting their correlations. Missing-at-random data, they show, should usually be ignored in single-element analysis, with a tailored fix for the multi-element case. They propose judging by effect size and an overlap coefficient, a probability of similarity, rather than p-values. Tests on synthetic and real proteomics data back the approach.
Catalog record
| CREATOR | Marie Chion, Arthur Leroy, Ilya Ioshikhes, Lun Hu, Ilya Ioshikhes, Ilya Ioshikhes |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 93 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |