Penn Engineers have developed PeptiVerse, an AI-powered platform that predicts key chemical and biological properties of peptides, which are strings of amino acids with significant medical potential, as demonstrated by the success of GLP-1 drugs, widely used for weight loss.
The researchers describe how they trained PeptiVerse using a wide range of data sets, allowing it to predict properties that can help determine whether a peptide is worth pursuing as a potential drug, such as the peptide’s likelihood of dissolving, entering cells, avoiding toxicity, and lasting long enough in the body to have an effect.
According to the report, tools for predicting such properties exist, but they often focus on a narrower set of traits or only one kind of peptide, whereas PeptiVerse brings many of those predictions together in one open-source, easily accessible platform.
How PeptiVerse Works
PeptiVerse allows users to evaluate both ordinary peptides and chemically modified versions designed to work better as drugs, making it a valuable tool for researchers.
“Peptide drugs have enormous potential, but binding to the right target is only one part of what makes a molecule useful,” says Pranam Chatterjee, senior author of the new study.
Chatterjee notes that in drug discovery, one of the worst outcomes is finding out too late that a promising molecule cannot actually 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.
Building PeptiVerse
To build PeptiVerse, the team first had to gather data that had been scattered across separate studies, which required synthesizing the data sets and standardizing the data for each property.
“Each data set came from a different type of experiment,” says Sophia Vincoff, a doctoral student and co-author of the study. “We had to understand what each experiment was measuring, standardize the data for each property and organize it so machine-learning models could learn from it and be tested fairly.”
The team compared many model architectures to find the best approach for each prediction task, rather than assuming that one model works best for every task.
Related Post: SAE Media to Host AI Drug Discovery Conference
Using PeptiVerse
PeptiVerse includes a web interface that allows users to simply type in a peptide sequence, select properties, and receive predictions through a visual dashboard, making it especially useful for experimental researchers.
“I come from a biology perspective, and I’m a very visual person,” says Yinuo Zhang, a postdoctoral researcher and the paper’s first author. “We wanted researchers to be able to see what was going on, not just download a Python package.”
The platform also lets users view the data used to train the models, helping them understand how their own peptide candidates compare with molecules that have already been experimentally characterized.
One immediate use of PeptiVerse is as a screening tool, allowing researchers to quickly assess which molecules appear most promising before moving into the lab.
Future Developments
PeptiVerse could also play a more ambitious role in AI-driven drug discovery, as it can be paired with generative AI tools that propose entirely new peptides, guiding the design of peptides with desired therapeutic features.
Chatterjee notes that the property predictions become part of the search itself. Instead of evaluating peptides only after they are generated, they can use those predictions to guide generative AI models toward molecules with the characteristics they want from the very beginning.
The team designed PeptiVerse to keep growing, as more peptide data becomes available, the platform can be updated with new data sets, improved models, and additional properties, which will help improve existing predictions and add new ones.
Since PeptiVerse is open source, researchers can adapt it for their own purposes, and companies developing peptide therapeutics could also use the framework with their own internal data sets, creating versions of PeptiVerse tailored to the kinds of molecules they are trying to develop.
Chatterjee says that the universe of possible peptides is too large for any one lab to map on its own. They 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.
