Bayesian Hierarchical Modeling for Label-Free Proteomics
Estimate measurement uncertainty
Function for Fitting the Mean-Variance Gamma Regression Models
Fit Latent Gamma Mixture Regression
Sample the Posterior of the data and decision model and generate point...
Latent Gamma Regression Model
The 'baldur' package.
Calculate the Mean-Variance trend
Baldur's empirical Bayes Prior For The Mean In Conditions
Estimate Gamma hyperparameters for sigma
Visualization of LGMR models
Pipe operator
Function for plotting the mean-variance gamma regressions
Function for plotting the gamma regression for the mean-variance trend
Plot the trend between the log fold-change and sigma, coloring signifi...
Plot the -log10(err) against the log fold-change
Normalize data to a pseudo-reference
Baldur's weakly informative prior for the mean in conditions
Statistical decision in proteomics data using a hierarchical Bayesian model. There are two regression models for describing the mean-variance trend, a gamma regression or a latent gamma mixture regression. The regression model is then used as an Empirical Bayes estimator for the prior on the variance in a peptide. Further, it assumes that each measurement has an uncertainty (increased variance) associated with it that is also inferred. Finally, it tries to estimate the posterior distribution (by Hamiltonian Monte Carlo) for the differences in means for each peptide in the data. Once the posterior is inferred, it integrates the tails to estimate the probability of error from which a statistical decision can be made. See Berg and Popescu for details (<doi:10.1016/j.mcpro.2023.100658>).