Leiden researchers, led by Sebastian Pomplun, have developed a new method to screen hundreds of thousands of molecules for drug discovery, using mass spectrometry instead of DNA tags. “We wanted to make drug discovery faster and more accessible,” said the researcher.
Finding a new medicine often starts with finding a molecule that binds to the right protein—a process that can take years and cost millions. Now, a team of researchers in Leiden has developed a faster and more flexible way to search for promising drug candidates, without the need for DNA barcodes. The findings are published in Nature Communications.
“In drug discovery, you usually start with a huge collection of molecules and hope that one of them sticks to your target protein,” explains Pomplun. “Traditionally, pharmaceutical companies test these molecules one by one in enormous robotic facilities, so-called high throughput screening. It’s effective, but it’s also incredibly expensive and slow.”
What if we could do the same kind of screening, but without the DNA?
Over the past decade, many labs have turned to DNA-encoded libraries (DELs). In this method, each small molecule is tagged with a short piece of DNA that acts like a barcode, recording what the molecule looks like. If a molecule binds to a target protein, researchers can simply read the DNA to find out which one it was.
“It’s a brilliant technology,” says Pomplun, “but it also has some big drawbacks.” The bulky DNA tag can block molecules from binding properly, especially to proteins that interact with DNA or RNA—and it limits the kinds of chemical reactions that can be used. “We thought, what if we could do the same kind of screening, but without the DNA at all?”
The ones that stick to the target protein are isolated and analyzed
Their new method replaces the barcode with mass spectrometry, a technique that can detect and identify tiny molecules based on their weight and how they break into fragments. The team designed chemical libraries with hundreds of thousands of compounds, each built from slightly different combinations of molecular building blocks. When these compounds are mixed with a target protein, the ones that stick are isolated and analyzed by mass spectrometry.
“It’s like each molecule leaves behind a unique fingerprint,” Pomplun explains. “Even if two compounds have the same mass, the way they fragment tells us which is which.”
A library of a half-million compounds in just a few days
Working with computational experts from the University of Jena, the Leiden team also developed software to interpret the complex spectra and match them to the right molecular structures. “That was a big step, turning all those signals into something we can actually use.”
The first results are promising: the method successfully identified nanomolar binders, or very strong hits, for cancer-related proteins, including targets that DNA-based methods can’t handle. “What excites me most is the speed,” says the researcher. “We can make a library of half a million compounds in just a few days.”
Making drug discovery more accessible for companies and academia
In the future, the team hopes their “self-encoded” approach will make drug discovery more accessible beyond big pharmaceutical companies. “With our method, academic labs could also more easily take part in early-stage drug discovery,” Pomplun says. “It’s faster, simpler, and opens the door to exploring new kinds of molecules that we couldn’t study before.”
More information:
Edith van der Nol et al, Barcode-free hit discovery from massive libraries enabled by automated small molecule structure annotation, Nature Communications (2025). DOI: 10.1038/s41467-025-65282-1
Provided by
Leiden University
The content is provided for information purposes only.