Using this information, the model can then tell us the probability of a drug-protein interaction that we did not previously have in the database, as the algorithms can efficiently analyse large ...