Dr. Valerie Mizrahi is a tuberculosis biologist at the University of Cape Town and the former director of its research institute. Her work runs from tuberculosis (TB) drug discovery through to the broader biology of Mycobacterium tuberculosis. Dr. Mizrahi is also a member of the TB Drug Accelerator, the world’s largest TB drug discovery consortium, convened by the Gates Foundation.
TB biology is complex almost everywhere you look, and the literature does not resolve that complexity so much as record it. Older papers do not cite the newer ones, newer papers do not always account for the older ones, and the finding that settles a question is often a passing remark rather than a headline. After less than a month with access, EMET is open on her laptop whenever the machine is on.
"As an enabler and as a time saver for me, it has transformed my work day."
A preprint that no search tool returned
TB literature tends to hide its most important details. "This literature of our field often relies on cryptic comments buried somewhere in a paper," Dr. Mizrahi said. So she tested EMET the way she tests a colleague. She asked a question, saw that two specific papers were missing from the answer, and said so.
EMET went back, ran a corpus search, retrieved the passage and returned a summary with the correct annotation.
Referencing is the thing she watches for.
"To me, the big thing you've got as a plus is the referencing."
Expression data scattered across a dozen repositories
Dr. Mizrahi is working on a project focused on differential gene expression in mycobacteria across many conditions. The RNA-seq and microarray data sit deposited across public repositories, and pulling out the genes of interest is, in her words, extraordinarily time consuming. She asked EMET under what conditions three specific genes had been shown to be differentially regulated.
"It searched all of these databases and gave me this beautiful summary."
When asked how much time EMET saves her, she said at least double. But the time is not the main thing:
"More than the saving of time for me is the ease with which I can get access to much more complex information... What does a scientist even do with 60 references?”
The scientist stays in the driving seat
Dr. Mizrahi checks EMET's work. She reads the reports carefully given she is an expert in the field. She has challenged EMET with the occasional simplistic conclusion in a generated figure, and digs deep into responses, just like a scientist would challenge a colleague.
"I'm not reliant on this being the ground truth or the gospel - that’s just the scientist in me."
"It reinforced my view that the power of these kinds of tools is best when coupled to people who know enough to query, to pick up inconsistencies, and to have an intelligent and credible back and forth."
That is the intended relationship. EMET is a co-scientist, not an oracle, and it is at its most useful in the hands of someone who knows where the buried detail should be. What Dr. Mizrahi’s experience shows is what an agentic workbench gives a lab when the person driving it is an expert: not answers to accept, but a much faster route to the evidence, with every citation traceable back to the paper it came from.
If you are an academic scientist looking for access to EMET be sure to sign up using the following link: https://emet.benchsci.com/
Want to explore what’s possible?
Try EMET
