J Proteome Res. 2022 Jun 10. doi: 10.1021/acs.jproteome.2c00161. Online ahead of print.
ABSTRACT
In their recent review ( J. Proteome Res. 2022, 21 (4), 849-864), Crook et al. diligently discuss the basics (and less basics) of Bayesian modeling, survey its various applications to proteomics, and highlight its potential for the improvement of computational proteomic tools. Despite its interest and comprehensiveness on these aspects, the pitfalls and risks of Bayesian approaches are hardly introduced to proteomic investigators. Among them, one is sufficiently important to be brought to attention: namely, the possibility that priors introduced at an early stage of the computational investigations detrimentally influence the final statistical significance.
PMID:35687673 | DOI:10.1021/acs.jproteome.2c00161
