The New Age of Seeing: What Science Finds When It Looks Again

In a Tanzanian museum basement, a specimen described by scientists more than 60 years ago sat waiting. When researchers at the University of Bristol finally unlocked its case, they found something unexpected: a herbivorous dicynodont named Dinodontosaurus isiyavamanda, dating back 240 million years to the Triassic. The discovery didn't just add a species to the catalog — it reshapes the timeline of early dinosaur evolution and reminds us that science often rewards those who look twice.

That same spirit of looking twice ran through laboratories this month, from Miami to the moon.

Searching by shape, not sequence

At Sylvester Comprehensive Cancer Center in Miami, Daniel G. Isom and his team asked a question that would have seemed impossible a decade ago. Using AI to compare the three-dimensional shapes of more than 214 million predicted proteins, they searched beyond genetic sequences for hidden members of the GPCR family — proteins that help cells sense and respond to extracellular signals.

"For decades, we have largely explored protein biology using sequence as our guide," said Isom. "What we found suggests there is another layer of biology that we've been missing."

The discovery, published in Nature, reveals a new way to study how cells communicate, survive stress and possibly contribute to disease. The AI didn't replace experimental science — it teamed up with it.

Nature's recyclers, co-opted

Meanwhile, in Denmark, Rosie Graham of Aarhus University took inspiration from a different kind of expert: the microbes already breaking down our garbage. Her team engineered an enzyme called CCPUR1, based on a bacterium found in compost, that can degrade polyurethane — the material in shoe foams, mattresses and kitchen sponges. Published in Chem Catalysis, the work targets roughly 22 million tons of polyurethane produced globally every year.

"We have flooded the environment with so much plastic of different kinds," Graham said. "But there are lots of organisms out there that are already using these materials as a source of energy."

Machines learning to see

The tools of discovery are themselves evolving. NASA, in collaboration with IBM Research, has released the NASA-IBM Lunar Foundation Model — one of the first open-source AI models built for lunar science, trained on 17 years of data from the Lunar Reconnaissance Orbiter. Hosted on Hugging Face for anyone to use, it helps researchers map ice, craters and volcanic features like Mons Rümker faster than ever before.

"NASA has spent decades building an extraordinary scientific record of the moon, but collecting data is only part of the job," said Kevin Murphy, NASA's chief science data officer.

The same pattern repeats across fields: at the Eastern Institute of Technology in Ningbo, researchers developed a graph-based approach that extracts interpretable physics equations directly from solid-material data, outperforming empirical models while preserving explicit mathematical form. In Prague and New York, Štěpán Jeřábek and colleagues at IOCB Prague and Columbia University shed new light on how early human embryos repair DNA damage — critical knowledge for the future safety of gene editing.

And the practical wins keep mounting. A team including USC researchers found that older adults who maintained higher physical function were less likely to experience hearing loss, challenging the assumption that cardiovascular health drove all age-related declines. Meanwhile, an international team including CIIMAR unveiled a genomic tool that could make Japanese clams and good clams more resistant to disease and climate change — a lifeline for coastal communities whose aquaculture has been declining.

The thread running through

Look closely enough at these eight studies and one theme emerges: we are getting better at seeing. Better at finding hidden proteins, hidden species, hidden laws of physics. AI is not replacing the human spark of curiosity — it's amplifying it, freeing scientists to ask bolder questions.

The next breakthrough may be waiting in a museum drawer, a compost heap, or 214 million protein structures you haven't bothered to compare yet. The only requirement is the willingness to look again.