A deep language model to predict metabolic network equilibria

François Charton, Amaury Hayat, Sean T. McQuade +2

We show that deep learning models, and especially architectures like the Transformer, originally intended for natural language, can be trained on randomly generated datasets to pre…