Why only dimmers (2 amino acids) are used as the key feature to train both the positive and negative data sets during gene predication?

Similar questions and discussions