ProteinPNet: Prototypical Part Networks for Concept Learning in Spatial Proteomics

TL;DR

ProteinPNet introduces a prototype-based learning framework that identifies interpretable spatial motifs in tumor microenvironments from proteomics data, aiding mechanistic understanding in spatial omics.

Contribution

It is the first to directly learn discriminative, interpretable spatial prototypes for TME analysis using supervised training, advancing spatial biomarker discovery.

Findings

Successfully identifies biologically meaningful prototypes in lung cancer data.

Prototypes reveal differences in immune infiltration and tissue structure.

Validated on synthetic datasets with ground truth motifs.

Abstract

Understanding the spatial architecture of the tumor microenvironment (TME) is critical to advance precision oncology. We present ProteinPNet, a novel framework based on prototypical part networks that discovers TME motifs from spatial proteomics data. Unlike traditional post-hoc explanability models, ProteinPNet directly learns discriminative, interpretable, faithful spatial prototypes through supervised training. We validate our approach on synthetic datasets with ground truth motifs, and further test it on a real-world lung cancer spatial proteomics dataset. ProteinPNet consistently identifies biologically meaningful prototypes aligned with different tumor subtypes. Through graphical and morphological analyses, we show that these prototypes capture interpretable features pointing to differences in immune infiltration and tissue modularity. Our results highlight the potential of prototype-based learning to reveal interpretable spatial biomarkers within the TME, with implications for mechanistic discovery in spatial omics.