From harnessing AI to integrating complex workflows, biotech leader Cytiva doesn't just supply tools for drug discovery -- it anticipates the industry's future challenges and develops solutions before its partners even know they need them
Flying on a plane out of Boston, Paul Belcher broke the habit of a lifetime. "I'm an introvert; I do not like people talking to me on aeroplanes," he says. But the man sitting next to him started chatting about his son, who was being treated for a rare genetic disease at Boston Children's Hospital."He was telling me about this new therapy that his son had and the hope it had given. And I knew the therapy, because I had worked to develop the analytical test to make sure it was safe and efficacious," says Belcher. It was a moment that brought home the human impact of his work. "I can retire happy with that," he says.
The test Belcher helped to develop, known as a QC release assay, is a vital part of the painstaking effort, often unseen, of transforming the scientific discoveries of today into the live-saving treatments of tomorrow.
It's a Herculean task. A decade ago, only 1 in 10,000 potential drugs developed in the lab made it into the clinic. Today, those odds may not feel any better as drug discovery becomes increasingly complex. Companies must design methods, or workflows, that weed out likely failures as early as possible in the drug discovery process and then take potential successes through development and production.
Biopharma experience
Belcher, a seasoned drug discovery strategist and business leader at the global biotechnology company Cytiva, and his colleagues, help biopharmaceutical companies tackle these challenges by supplying the equipment they need as well as the expertise to discover and manufacture advanced medicines.
This support goes well beyond improving a company's existing processes. Cytiva's breadth and depth of expertise gives its scientific teams the insights they need to unearth gems of their own, often creating innovations before their clients even know they need them.
"It encompasses the entire space of pharmaceuticals that are currently in discovery, both small-molecules and biologics, but also cell-based and viral-based therapeutics," says Belcher's colleague, Martin Teichert, who heads a product-innovation team that works closely with customers via collaborative efforts.
Cytiva was formed in 2020 when global science and technology leader Danaher acquired the life science business of medical technology company GE Healthcare. Its portfolio includes ÄKTA chromatography systems and the Biacore SPR (surface plasmon resonance) platform, which uses light to understand how molecules -- such as drugs and proteins -- interact, without the need for special tags or dyes that could interfere with their behaviour.
These are just some of Cytiva's many offerings. "A lot of our customers in the biotherapeutic space are trying to compress workflows, so we supply those tools," says Teichert. "But it's not just something you order from a catalogue and use. Many of our users want to give feedback and want to give input to further develop the technologies."
This goes beyond the typical company-customer relationship. "They are ultimately wanting to get those new therapies to market faster, to de-risk them earlier," says Belcher. "And I think that's where they look to us as partners to innovate to solve those kinds of problems."
Thanks to its close relationship with its clients, and the detailed customer research it carries out, Cytiva has identified five key needs that are currently unmet across the industry. The first is the need to compress timelines -- screen more molecules, and faster. This demands more robust, automated assays. Second, targets are becoming more complex, meaning samples become more precious. This demands the miniaturisation of technologies to do more with less sample. Third, the need to maximise the amount of information gleaned from each sample, to allow faster decision-making and to identify potential molecule liabilities and prioritise good leads sooner. Fourth is the desire to "hyphenate" technologies by combining them into a single unit op to reduce the number of workflow steps, saving time.
And finally, there is artificial intelligence and how it could be used to design better, safer, more effective therapies. Much of Cytiva's effort in this arena has focused on ensuring that the data produced by different workflows is standardised and pooled in "data lakes", which can be explored with AI. This is key, as the nature of the data is rapidly changing. "In the past, it's been more binary. Now we see much more information-rich data," says Teichert.
These kinds of insights allow Cytiva to develop solutions to problems that customers aren't yet aware of. About five years ago, Teichert and Belcher commercialised one of Cytiva's first AI-based products. It automated the analysis of large datasets, saving users 90 per cent of the time they would otherwise spend on this task.
Fresh insight
Ironically, this was not something their clients had realised was an issue. "They just accepted that's how it was," says Belcher. Company scientists are focused on discovering drugs rather than on the processes to do so, he adds, which is why having a fresh, external pair of eyes can help. "I guess that's what often breakthrough innovation looks like," says Teichert.
Another major bottleneck in drug discovery is ensuring that a drug hits a specific "target", a protein associated with a particular disease. If the drug doesn't inhibit, enhance or modulate the function of its target, it won't work, and lack of efficacy is one of the main reasons a drug fails in clinical trials. "The whole Alzheimer's space is a litany of examples of that," says Belcher.
To minimise this risk, companies are trying to improve target identification and validation, including testing different protein constructs to find ones that are best suited to the screening or analytical tools to identify new molecular entities that engage the target. But expressing and testing different constructs usually takes months and involves engineering cells to produce the protein fragments of interest.
Biotech company Nuclera is transforming this field by creating a benchtop machine that expresses and purifies protein constructs of interest designed using AlphaFold AI, without the need for cells and in a fraction of the usual time.
But producing proteins is only the first step. You next need to study their characteristics and pick optimal construct(s) to use in large-scale experiments to find molecules that bind to it.
In collaboration with Nuclera, Cytiva has demonstrated that its Biacore system can rapidly and accurately perform this step. And integration of the two technologies into one workflow can reduce the time needed to produce and characterise proteins from weeks or months to days. "It shows how these technologies can work together and really align with where most of the pharma companies are going right now, the utilisation of more AI in the drug discovery process," says Teichert.
Future drug discovery is likely to lean even more on AI but will always need a human touch. And it has the power to change people's lives -- as Belcher discovered on his Boston flight. "It shows I should talk to people more on planes," he quips. "I may learn something more interesting."
Find out more at: www.cytivalifesciences.com/solutions/protein-research