Sponsored content brought to you by
GEN: Why was Applied BioMath founded?
Dr. Burke: While at MIT and Harvard Medical School, and working as a consultant for pharma firms and large and small biotechs, I repeatedly saw how people were not using hardcore mechanistic mathematical modeling and biology with hardcore mathematics and high performance computing to generate data to de-risk important questions.
Pockets of people and groups at large pharma firms were employing mathematical modeling, but it wasn’t being done systematically at all companies. We started Applied BioMath to help our clients make better decisions and de-risk their projects, identify early failures, and save a lot of money and opportunity time. Our approach involves aspects of engineering that we want to bring to pharma firms to help them get their products to the clinic faster.
GEN: How has this translated into certain techniques that you use to revolutionize drug invention?
Dr. Burke: We bring the right mathematics and approaches to the fore as a function of our partners, their questions, data, and timelines. We consider biology just as critical as mathematics, and we utilize high-performance computing so we can do thousands to hundreds of thousands of simulations quickly.
All our project teams have at least one modeler, one biologist, and one experienced leader. We also typically have mentors that provide advice. Internally, we vet what we’re doing to the nth degree. We work closely and well with our partners, and I believe that this has contributed to our high repeat business rate—at least 80%.
GEN: How do you help companies incorporate all their data to make better clinical decisions, like first-in-human dose predictions?
Dr. Burke: I would take it a step further than that. We want to help identify the failures so they occur as early in the pipeline as possible. Ideally, you want to fail the day after a project starts, rather than after Phase I, II, or III. If you’re going to fail, fail early and hard. We work with our partners and do a literature search, and then build the models and analyze them. We learn as much as possible about the disease biology and how the drug works mechanistically, as quickly as possible.
For example, with a biologic, there’s usually a Kon and a Koff. Maybe it’s a bispecific that’s a tetravalent binder. You’re going to have two different Kon’s and Koff’s, potentially, for each epitope in each arm. And then you’re hitting certain targets, and those targets differ. There are certain cells that maybe you do want to hit, and there’s others you don’t, which impacts therapeutic index. So, even before our partners start generating any data, we can talk about how hard this project is going to be to execute or to even develop a drug. Then, as the winners progress, we help them identify what properties make their candidates best-in-class, thus optimizing lead generation and prioritizing experiments so you can get to IND faster, hopefully as a best-in-class and first-in-class therapy. We can update the model with human data to provide better predictions and guidance in the clinic.
GEN: How might Applied BioMath contribute to research on gene and cell therapies?
Dr. Burke: There are so many unknowns to designing a cell and gene therapy. How would one perform allometric scaling for these therapies? Our experience shows this is not the best approach. Once you use mechanistic modeling in these projects, you can say to yourself, “You know what? These parameters don’t matter as much as others.” For example, sometimes target expression is a sensitive parameter, sometimes not. Sometimes protein synthesis is a sensitive parameter, other times not. How critical is the expansion phase or half-life in your cell therapy? From sensitivity analysis, we identify these parameters. Just like other programs, we can really use the biology in a directed way because of modeling. Because we’re using math and biology dynamically with high-performance computing, hopefully you can design that best-in-class drug even though you’ve never done it before.
Learn more about Applied BioMath at appliedbiomath.com