A semi-supervised Bayesian approach for simultaneous protein sub-cellular localisation assignment and novelty detection

Abstract

The cell is compartmentalised into complex micro-environments allowing an array of specialised biological processes to be carried out in synchrony. Determining a protein’s sub-cellular localisation to one or more of these compartments can therefore be a first step in determining its function. High-throughput and high-accuracy mass spectrometry-based sub-cellular proteomic methods can now shed light on the localisation of thousands of proteins at once. Machine learning algorithms are then typically employed to make protein-organelle assignments. However, these algorithms are limited by insufficient and incomplete annotation. We propose a semi-supervised Bayesian approach to novelty detection, allowing the discovery of additional, previously unannotated sub-cellular niches. Inference in our model is performed in a Bayesian framework, allowing us to quantify uncertainty in the allocation of proteins to new sub-cellular niches, as well as in the number of newly discovered compartments. We apply our approach across 10 mass spectrometry based spatial proteomic datasets, representing a diverse range of experimental protocols. Application of our approach to hyperLOPIT datasets validates its utility by recovering enrichment with chromatin-associated proteins without annotation and uncovers sub-nuclear compartmentalisation which was not identified in the original analysis. Moreover, using sub-cellular proteomics data from Saccharomyces cerevisiae, we uncover a novel group of proteins trafficking from the ER to the early Golgi apparatus. Overall, we demonstrate the potential for novelty detection to yield biologically relevant niches that are missed by current approaches.

A Method to Estimate the Distribution of Proteins across Multiple Compartments Using Data from Quantitative Proteomics Subcellular Fractionation Experiments

“…If one or more of the proteins in such a group is known to reside in a compartment that is not represented in the marker protein set, it would be possible to use this knowledge to create an additional reference compartment for the CPA procedure. Alternatively, a Bayesian approach may be helpful in detecting previously unassigned subcellular compartments . Given high enough resolution fractionation methods, it may also be possible to further subdivide the major compartments in our current classification scheme.…”

Semi-supervised nonparametric Bayesian modelling of spatial proteomics

“…As described in the Introduction, our data includes a set of marker proteins whose subcellular niche localisations are known a priori. Thus, K is known from the outset, since we assume that we have, at least, one marker protein localising to each subcellular niche; that is, we assume that all classes are represented among our labelled data (see Crook et al (2020), for a relaxation of this assumption).…”

Challenges and Opportunities for Bayesian Statistics in Proteomics

“…One approach is spatial subcellular proteomics , where proteins are localized to their subcellular niche using mass-spectrometry data. Bayesian approaches have been developed for biochemical fractionation-based subcellular proteomics. ,,,, Crook et al ,, demonstrated that Bayesian modeling can quantify uncertainty in protein subcellular localization and identify cases where this may correspond to multilocalizing proteins. Crook et al showed that even a Bayesian point estimate may overlook these cases, and more information is obtained by examining the full posterior distribution.…”