Penn Engineers announced on July 21 the development of PeptiVerse, an artificial intelligence-powered platform designed to predict key chemical and biological properties of peptides. The tool aims to help researchers determine whether a peptide is suitable for further drug development by evaluating traits such as solubility, cell entry, toxicity avoidance, and stability in the body.
According to a study published in Nature Communications, the research team trained PeptiVerse using diverse data sets. This approach enables the platform to assess both standard peptides and chemically modified versions intended for therapeutic use. "Peptide drugs have enormous potential, but binding to the right target is only one part of what makes a molecule useful," said Pranam Chatterjee, Africk-Lesley Distinguished Scholar of Innovation in Engineering and senior author of the study.
Chatterjee said that one major challenge in drug discovery is realizing too late that a promising molecule cannot become a medicine. "PeptiVerse gives researchers a way to check many of those make-or-break properties earlier, before they invest the time and resources required to synthesize and test candidate drugs." To build PeptiVerse, researchers gathered experimental data from various studies on peptide dissolution, cell entry capabilities, resistance to unwanted protein buildup, toxicity avoidance, and longevity.
Sophia Vincoff, doctoral student and co-author of the study, said synthesizing these data sets involved standardizing information from different experiments so machine-learning models could learn effectively. The team compared multiple model architectures for each prediction task rather than relying on one universal approach. Yinuo Zhang, postdoctoral researcher and first author of the paper, said, "We wanted PeptiVerse to function like a toolkit... Instead of having each predictor live separately, PeptiVerse puts them together in a platform that can grow as new data and models become available."
The web-based interface allows users without programming experience—such as biologists—to input peptide sequences directly into an interactive dashboard for predictions. Zhang added, "We wanted researchers to be able to see what was going on... The interface makes PeptiVerse something people can interact with directly." Users can also view training data behind predictions for greater transparency.
Chatterjee concluded that open-source access will allow other academic labs or companies developing peptide therapeutics to adapt or expand PeptiVerse with additional data or custom predictors: "We built PeptiVerse so that other researchers can help expand the map, adding new data and models that accelerate the search for new and better peptide drugs."
