A method for finding high-affinity antibody therapeutics that is based on artificial intelligence (AI) has been created by researchers at the University of California San Diego School of Medicine.
Researchers utilised the method in the study, which was released on January 28, 2023 in Nature Communications, to discover a novel antibody that binds a significant cancer target 17-fold more tightly than an antibody medication currently on the market. According to the scientists, the pipeline may hasten the development of new medications to treat conditions like COVID-19 and rheumatoid arthritis as well as cancer.
A successful drug requires an antibody to bind firmly to its target. Researchers usually start with a known antibody amino acid sequence in order to locate these antibodies, and then they use bacterial or yeast cells to make a series of novel antibodies with variations on that sequence. Then, the mutants’ capacity to bind the target antigen is assessed. The best-performing subset of antibodies is then put through another round of modifications and assessments, and this cycle is repeated until a short list of tightly-binding finalists is established.
Many of the produced antibodies still show poor results in clinical trials despite this time-consuming and expensive process. To expedite and streamline these efforts, UC San Diego researchers created a cutting-edge machine learning algorithm for the current study.
Similar to other approaches, this one begins with creating an initial library of roughly 500,000 potential antibody sequences and screening them for their affinity to a particular protein target. The dataset is fed into a Bayesian neural network, which may interpret the data and utilise it to forecast the binding affinity of new sequences, rather than repeatedly repeating this procedure.
Wei Wang, Ph.D., senior author and professor of cellular and molecular medicine at the UC San Diego School of Medicine, noted that successive rounds of sequence mutation and selection can be completed swiftly and effectively on a computer as opposed to in a lab.
Their AI model’s capacity to report the accuracy of each forecast is one of its key advantages. We can rank the antibodies and choose which ones to prioritise in medication development since, unlike many AI techniques, our model can truly communicate how confident it is in each of its predictions, according to Wang.
Jonathan Parkinson, Ph.D., and Ryan Hard, Ph.D., project scientists and co-first authors of the study, set out to develop an antibody against programmed death ligand 1 (PD-L1), a protein that is highly expressed in cancer and the target of a number of commercially available anti-cancer drugs, in order to validate the pipeline. By employing this method, scientists discovered a new antibody that bound to PD-L1 17 times more effectively than atezolizumab (Tecentriq), the wild-type antibody that the U.S. Food and Drug Administration has licenced for use in clinical trials.
This method is currently being used by the researchers to find potential antibodies against additional antigens, such as SARS-CoV-2. They are also creating new AI models that examine amino acid sequences for other antibody qualities including stability, solubility, and selectivity that are crucial for the outcome of clinical trials.
By combining these AI techniques, Wang stated, “scientists may be able to execute a greater proportion of their antibody discovery work on a computer instead of at the bench, perhaps leading to a faster and less prone to failure discovery process.” “This pipeline has a plethora of applications, and these results are truly only the tip of the iceberg.”
