Orbitrap noise structure and method for noise unbiased multivariate analysis

Orbitrap mass spectrometry is widely used in the life-sciences. However, like all mass spectrometers, non-uniform (heteroscedastic) noise introduces bias in multivariate analysis complicating data interpretation. Here, we study the noise structure of an Orbitrap mass analyser integrated into a secondary ion mass spectrometer (OrbiSIMS). Using a stable primary ion beam to provide a well-controlled source of ions from a silver sample, we find that noise has three characteristic regimes: at low signals the Orbitrap detector noise and a censoring algorithm dominates; at intermediate signals counting noise specific to the ion emission process is most significant; and at high signals additional sources of measurement variation become important. Using this understanding, we developed a generative model for Orbitrap data that accounts for the noise distribution and introduce a scaling method, termed WSoR, to reduce the effects of noise bias in multivariate analysis. We compare WSoR performance with no-scaling and existing scaling methods for three biological imaging data sets including drosophila central nervous system, mouse testis and a desorption electrospray ionisation (DESI) image of a rat liver. WSoR consistently performed best at discriminating chemical information from noise. The performance of the other methods varied on a case-by-case basis, complicating the analysis.