A third of his time back, and a protein he can finally detect

Dr. Fasih Rahman is a muscle physiology researcher at the University of Guelph. His work sits under muscle remodeling, the changes in both the quality and the quantity of muscle, across preclinical models of physical inactivity, cancer, and aging. His focus has recently shifted toward the neuromuscular junction, and his data has started pointing him toward extracellular matrix remodeling, a field that is new to him.

He did not come to EMET cold. He had already been using other AI tools for literature review and had tried building figures with it, so he had a basis for comparison. He was also not running test prompts. He was running live experiments, shaping grant ideas, and fighting an experimental protocol that remained challenging for several weeks.

"Absolutely, it is probably saving a third of my time."

A buried methods detail from EMET that rescued an assay

Dr. Rahman could not detect MuSK, a protein at the neuromuscular junction in his samples. His lysis buffer had Triton X and sodium deoxycholate in it, which (according to published protocols) he expected to be enough. It was not.

So he gave the problem to EMET. EMET read the literature and came back with the one variable he had missed: the addition of 0.1% SDS. Buffer composition like this is buried in methods sections, not abstracts. EMET reads the full text of papers, including closed access literature, which is where that detail was sitting.

"It went through the literature. It caught one thing that I missed, and that was the addition of 0.1% SDS. Added it in, and sure enough, I can detect it now."

He rehomogenized his samples and immunoprecipitated. The phosphorylation status held, everything he expected to see was there, and the blots were clean. After several weeks of troubleshooting, a small methodological detail identified in the literature helped improve the assay.

Since then he has used EMET to synthesize known protocols before he runs them.

The small papers that carry the exact finding

For his ECM remodeling review work, Dr. Rahman asked EMET for the literature linking specific processes in muscle in the context of cancer. He rates both tools he uses as competent at orienting him in a field. The gap shows up at the edges.

"I do find that EMET works a little bit better at finding some more of those niche papers that have helped me orient myself to what is happening in the field."

The example he gave was a link between TNF alpha, ECM remodeling, and MuSK signaling that he had found that morning. He also checks EMET's citations against the papers themselves.

"Those quotations that it provides for that paper, where explicitly they say, or they show, or they are hypothesizing a certain fact, it does a very good job at that. I have cross referenced this a number of times."

He no longer searches for papers manually. "I seldomly use Google alone for my searches. I find that using AI tools like this has significantly accelerated how quickly I can identify relevant literature and build my understanding of a field"

Dashboards he prints out and marks up by hand

Dr. Rahman builds dashboards in EMET for different projects and grant ideas, along with schematics and rough figure outlines for workflows such as his proteomics pipeline.

"For creating the dashboards, without a doubt it has done considerably better."

He also noticed that he stopped having to specify the format he wanted. With his other tool he has to ask explicitly for a table, in a particular layout, with particular columns. EMET infers it, so his prompts have gotten easier to develop.

The scientist stays in charge

None of this replaces his judgement, and he does not use it that way. He treats EMET's output as a starting point, then goes back to his own data and looks harder. He reads the lists it returns himself. He cross references the quotations it pulls. He prints the dashboards and marks them up by hand.

This experience shows where time savings can become meaningful. Dr. Rahman would have found a solution eventually, but identifying a relevant methodological detail that takes considerable time when it is buried in methods sections of papers. An agentic research workbench can help streamline that process by surfacing potentially useful information more efficiently and as a result allow researchers to spend more time testing and building those findings.

If you are an academic scientist looking for access to EMET be sure to sign up using the following link: https://emet.benchsci.com/