#AskAbsci: How does Absci address key failure points in the drug discovery process?
May 11, 2022
Question: How does Absci address key failure points in the drug discovery process?
Answer: Today’s conventional approach to drug discovery, development, and manufacturing is inefficient, time-consuming, and expensive. It involves a number of sequential steps, each of which relies on different technologies and presents different technical challenges. Optimizing at one step may have a negative impact on the next, or the next, which unfortunately may not be uncovered until much later in the process. For example, an early discovery team may discover a molecule that has the desired drug-like properties, but that is ultimately difficult to manufacture or suffers from other developability challenges. This leads to long timelines and high failure rates, even before drug candidates reach clinical testing. In addition, conventional approaches are adaptable only to discovering new drug candidates from a repertoire of sequences that already exist in a prebuilt library. If a sequence isn’t present in the library to start with, you won’t find it. Given that there are more possible variants of an antibody sequence than there are atoms in the known universe, these pre-built libraries only scratch the surface of the potential diversity that we would like to explore.
How can we do this better, faster, more efficiently?
We looked at the endgame, getting better drugs to patients – faster and with higher success rates – and asked how? We are building our technology to be that how.
Our Integrated Drug Creation PlatformTM unifies the drug discovery and cell line development processes in a single workflow. With our platform we screen for the best drug candidate while simultaneously creating the optimized cell line for scalable manufacturing. An additional benefit to our platform is that we generate proprietary and extremely useful data about protein binding affinities, manufacturability of sequences, and genetic determinants leading to a successful manufacturing cell line , and these data are a perfect substrate for AI modeling.
In fact, we have built out a world class AI team and we are developing and validating in the wet-lab, cutting-edge AI-powered in silico models that can be used to drive even better efficiency through computational drug design and lead optimization. We hope to use these approaches to accomplish a lot of the work that is conventionally done through laborious screening and trial and error. Our aim is to transform the protein drug discovery process from target to clinic, improve development of successful drugs, and ultimately get better medicines to patients.