CPP Knowledge Graph

Cell-penetrating peptides (CPPs) are short sequences characterized by their ability to translocate across cell membranes without causing significant damage. This property enables CPPs to function as molecular carriers, facilitating the intracellular delivery of bioactive cargoes. Despite their potential, the discovery of new CPPs relies on time-intensive experimental validation; in silico methodologies offer an alternative solution. However, existing tabular databases limit the extraction of relational biological context, which restricts the capacity of computational models to learn biologically meaningful relationships and reduces the performance of predictive algorithms. To the best of our knowledge, this is the first knowledge graph that integrates cell-penetrating peptides with their associated cellular uptake mechanisms, cell lines, cargo molecules, and subcellular delivery locations. Moreover, to establish a functional biological context, we linked the specific uptake mechanisms to regulatory genes and chemical inhibitors. To construct this resource, we collected 5,310 peptide sequences from CPPsite3, and we processed the annotations through a normalization pipeline and used a graph retrieval-augmented generation method (Graph RAG) for ontology mapping. We mapped annotations such as uptake mechanisms, subcellular locations, cell lines, and cargo types to existing Gene Ontology (GO), Cell Line Ontology (CLO), and Chemical Entities of Biological Interest (ChEBI) to ensure interoperability. We serialized the dataset as a Resource Description Framework (RDF) graph and employed the Semanticscience Integrated Ontology (SIO) as an upper-level ontology to further improve interoperability; researchers can access the source code and web interface at https://github.com/bio-ontology-research-group/Cell-penetrating-peptides and https://cppkg.bio2vec.net/.