pRoloc

pRoloc infers protein sub-cellular localization from quantitative mass spectrometry (MS) data to map spatial protein distributions and support analysis of post-translational regulatory mechanisms.

Key Features:

• Transfer learning integration: Implements a transfer learning classification framework that integrates heterogeneous auxiliary data such as immunofluorescence microscopy and protein annotations/sequences with MS data.
• Classification algorithms: Applies nearest-neighbour and support vector machine classifiers to assign proteins to sub-cellular compartments.
• Cross-species and multi-condition robustness: Handles complex datasets from different species and multiple experimental conditions to maintain generalization accuracy.
• Pattern recognition for localization: Leverages advanced pattern recognition techniques to pinpoint spatial protein distributions relevant to post-translational regulation.

Scientific Applications:

• High-throughput spatial proteomics: Analysis of MS-based spatial proteomics experiments to generate sub-cellular localization maps.
• Proteome-wide localization mapping: Precise localization of thousands of proteins under controlled experimental conditions.
• Regulatory mechanism investigation: Exploration of cellular functions and post-translational regulatory mechanisms via spatial distribution data.
• Discovery of novel localizations: Classification of previously unknown proteins, including applications to pluripotent mouse embryonic stem cells.

Methodology:

Uses a transfer learning classification framework to integrate immunofluorescence microscopy and protein annotations/sequences with quantitative MS data, applies nearest-neighbour or support vector machine classifiers to assign proteins to sub-cellular compartments, evaluates experimental datasets from multiple species, and validates results against high-resolution maps of the mouse stem cell proteome.