I was reading some papers about protein-protein interaction prediction using machine learning. I understood that using various algorithms, all the sequences are transformed into equal length vector. But I couldn't understood how these data are presented in training/test set. Because interaction happens between two proteins, how features of two proteins can be taken into account for a label like 'has interaction' and 'interaction'?

More Ruhshan Ahmed Abir's questions See All
Similar questions and discussions