LRP

LRP is an acronym used in several domains, most notably in biology and computer science. It may refer to lipoprotein receptor-related proteins, the leucine-responsive regulatory protein in bacteria, or Layer-wise Relevance Propagation in AI.

Lipoprotein receptor-related proteins are large transmembrane receptors in the LDL receptor family. Notable members include LRP1 and LRP2, also called megalin. They have extensive extracellular ligand-binding domains, a single transmembrane segment, and a cytoplasmic tail with endocytic signals. LRP1 and LRP2 mediate endocytosis and clearance of diverse ligands such as lipoproteins, proteases, and vitamin carriers. LRP5 and LRP6 function as co-receptors in the Wnt signaling pathway and influence development and bone metabolism. Dysfunctions or altered signaling through LRP receptors have been linked to cardiovascular disease and neurodegenerative processes, including effects on amyloid-beta clearance in the brain.

Leucine-responsive regulatory protein is a bacterial transcription factor of the Lrp/AsnC family. It senses intracellular leucine and forms oligomers that modulate DNA binding, acting as a global regulator of operons involved in amino acid synthesis, transport, and metabolism. Lrp can function as a repressor or activator and often coordinates with other regulators to adapt gene expression to nutrient availability.

Layer-wise Relevance Propagation is an explainable artificial intelligence technique that attributes a neural network’s decision to its input features. It backpropagates a relevance score from the output toward the input, using rules that conserve total relevance. Variants such as epsilon- and gamma-based rules address numerical stability and sensitivity to weights. LRP is applied to interpret models in vision and signal processing, but its behavior depends on architecture, preprocessing, and the chosen rules, and it remains an active area of evaluation.