If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.


Drug Discovery

Generate Biomedicines opens cryo-EM lab to feed data back into its protein design algorithms

Facility aims to characterize AI-desgined protein structures at scale

by Laurel Oldach
July 13, 2023 | A version of this story appeared in Volume 101, Issue 23


A man in front of a microscope twice as tall as he is and about as deep as a walk-in closet, façade open to show a tangle of tubes and parts.
Credit: Lara Woolfson/Studio Nouveau
Generate Biomedicines scientist Claudio Perez poses with a cryo-electron microscope.

Generate Biomedicines, which uses artificial intelligence for protein design, recently opened a cryo-electron microscopy (cryo-EM) laboratory in Andover, Massachusetts, that it says will accelerate its efforts to discover new protein-based therapeutics for oncology, immunology, and infectious diseases.

Cryo-EM enables scientists to determine the structure of a protein without first crystallizing it, and has revolutionized structural biology in recent years. “What no one’s been able to crack though, yet, is just how do you scale this up?” says Generate cofounder and chief technical officer Gevorg Grigoryan, who says he envisions a future where determining a protein structure is as routine as sequencing a gene.

To acquire structural data around the clock, Generate spent about $15 million on four automated microscopes from Thermo Fisher Scientific and JEOL and roughly another $6 million on lab renovation. The company uses automated lab systems to help standardize production of AI-designed protein samples from DNA synthesis through imaging and is developing its own automated microfluidic lab systems to improve sample preparation.

While several pharmaceutical companies use machine learning for de novo protein design, JEOL sales account manager Jens Breffke says Generate is unique in its workflow: producing those proteins and confirming that they are shaped as expected. Its scientists hope that rapidly generating high resolution structures will both help fine-tune potential therapeutics and improve its models’ performance.

According to Melanie Adams-Cioaba, general manager of the pharmaceutical cryo-EM segment at Thermo Fisher Scientific, Generate “sits at the forefront of a lot of where both computation and experimentation are pushing.”



This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.