Abstract: Provided herein are improved methods for training graph neural networks (GNNs) to predict the binding affinity of novel compounds based on data generated from DNA-encoded library (DEL) experiments. These methods include training the GNN to predict affinity for the target directly and then applying the predicted affinity to a model of the DEL experiment process to generate a predicted DEL read count. The predicted DEL read count can then be compared to an experimentally-observed DEL, read count to generate a loss value. The loss value can then be used to update the GNN as part of a GNN training process. The loss value can be augmented with simulated disynthon data generated from the predicted affinity.

Abstract: A system and method for separating permanent magnets from a rotor for repurposing. The method includes obtaining a first rotor including a plurality of permanent magnets located in a plurality of cavities arranged in the first rotor, heating the first rotor to a first temperature, removing the permanent magnets from a respective cavity by applying a force onto the permanent magnets at the first temperature, and repurposing the permanent magnets for a secondary use. The first temperature may be greater than a glass transition temperature and less than a melting temperature. The permanent magnets may be further processed for installation into a second rotor, the processing may include grinding the one or more permanent magnets to enable installation into the second rotor, analyzing a structural integrity of the permanent magnets to determine suitability for repurposing, and may also include separating the polymer material from a magnetic material.