Abstract
Therapeutic antibodies have emerged as the most prominent class of new drugs due to their high specificity and their ability to target several molecules at the same time, in the case of multi-specific bio-therapeutics. Once an antibody has been identified, an optimization phase follows that includes affinity maturation, often based on the structure of the interface between the target and the antibody.
Cryo-EM is quickly emerging as the most efficient method to obtain such structures. It provides a noisy tridimensional map of probability of presence of electrons. The number, nature and structure of antibodies have to be inferred from the map. Automated methods exist for solving the more general problem of finding any protein in a density, but have a limited accuracy and usually require additional inputs, high resolution maps, and a long running time.
We propose a method dedicated to finding antibodies in cryo-EM densities : CrAI. This machine learning method leverages the conserved structure of antibodies and a dedicated novel database to solve this problem. We show that our method performs favorably compared to other methods, alleviating aforementioned limitations. Running a prediction takes only a few seconds and requires nothing but the cryo-EM density.
Competing Interest Statement
The authors have declared no competing interest.
