DrugTar Algorithm

We developed "DrugTar", a novel deep learning framework that integrates pre-trained sequence embeddings and GO term for predicting druggability. The overview of this method is shown here:

DrugTar leverages the combined power of the ESM-2 embedding method and GO annotations within a deep neural network framework. Effective feature selection in biological systems with high-dimensional features is critical to improving classification accuracy and preventing overfitting. Therefore, the SVM feature selection algorithm was employed on the concatenation of ESM-2 embedding and GO terms in DrugTar. Subsequently, a deep neural network with three hidden layers was utilized to predict protein druggability. DrugTar achieved areas under the curve and precision-recall curve values above 0.90, outperforming state-of-the-art methods. In conclusion, DrugTar streamlines target discovery as a bottleneck in developing novel therapeutics. For more information, see DrugTar Improves Druggability Prediction by Integrating Large Language Models and Gene Ontologies.