An updated Bioconductor workflow for correlation profiling subcellular proteomics.

Background: Subcellular localisation is a determining factor of protein function. Mass spectrometry-based correlation profiling experiments facilitate the classification of protein subcellular localisation on a proteome-wide scale. In turn, static localisations can be compared across conditions to identify differential protein localisation events. Methods: Here, we provide a workflow for the processing and analysis of subcellular proteomics data derived from mass spectrometry-based correlation profiling experiments. This workflow utilises open-source R software packages from the Bioconductor project and provides extensive discussion of the key processing steps required to achieve high confidence protein localisation classifications and differential localisation predictions. The workflow is applicable to any correlation profiling data and supplementary code is provided to help users adapt the workflow to DDA and DIA data processed with different database softwares. Results: The workflow is divided into three sections. First we outline data processing using the QFeatures infrastructure to generate high quality protein correlation profiles. Next, protein subcellular localisation classification is carried out using machine learning. Finally, prediction of differential localisation events is covered for dynamic correlation profiling experiments. Conclusions: A comprehensive start-to-end workflow for correlation profiling subcellular proteomics experiments is presented.