Antiverse: Taking drug development from 18 months to 18 hours.

We first met Rowina whilst running a workshop at Imperial’s incubator on exploring the potential of personalised medicine. With a background in cell free synthesis systems for high yield protein expression she seemed perfect for DSV. We decided to invite her to join as a part time Exec Fellow to develop a game-changing startup; it turned out the ideal opportunity was awaiting!

Next came Murat: he applied directly after seeing the write up in New Scientist. Having worked as a lab automation programmer for 12 years he had initially intended to join as an Exec Fellow, but when his employer wouldn’t allow that he quit to join DSV as the only full time member of the team.

The final member of the team is Ben, a mechanical engineer by training who had previously taken adjustable-focus eyeglasses from concept to manufacture. He had taken a career break to pursue his long-standing interest in machine learning whilst raising his son.

The team didn’t work together, or on antibodies, right away. As was the DSV blueprint, they explored many ideas with many members of the cohort to find truly great need/team/solution matches. Ideas included distributed cloud labs using dormant equipment, an integrated synbio development environment, cell activity simulation, parasitic worms as a treatment, predicting the response of the microbiome to particular drugs and reducing the impact of cytokine storms during CAR-T cell therapy.

However, after around a month it was clear that they were all equally passionate about chipping away at the insane time and cost it takes to develop a new drug (6–10 years and $2.6bn). If you can make the process quicker and cheaper it vastly increases the chance of developing effective treatments.

The team undertook a sprint to examine the drug development processes and identified that antibodies are the fastest-growing drug type and already have a large market share. Moreover antibody discovery is still conducted in the lab, partly by screening antibodies but largely by exposing animals to the antigen, followed by an arduous process of extracting, filtering and scaling up quantities of the antibody, taking up to 18 months and at a cost of £30k–300k per candidate.

The idea of using machine learning and cell free protein synthesis to replace this process started to come together within the next couple of weeks. A very rough proof of concept showed that it could provide results that equaled those of traditional methods in just a few hours. With this data the team quickly found traction with major contract research organisations, the outsourcing agents which handle around $1bn of antibody discovery for large pharma companies.

The team recognised that it can be difficult to raise money for what is predominately software in the field of pharma (more on that madness in another post), but they were confident that they had made an important breakthrough, and through their deep technical expertise and with intensive customer development they forged a collaboration with a major pharma player, giving them market credibility.

The team are currently building out the core technology, working up partnerships and raising investment.

Join us at Deep Science Ventures to build a deep-tech venture that matters. Other enquiries to




We combine available scientific knowledge and founder-type scientists into high-impact ventures.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Mark Hammond

Mark Hammond

Founder at @deepsciventures creating a new paradigm for applied science. Ex-neuropharmacologist & AI researcher.

More from Medium

Interview Of The Week: Christian Lindener

Energy usage in MSMEs — An outlook

What’s Your (Learning) Agenda?

Product Thinking- Myth vs Reality