I am trying to make dataset in order to predict subcellular localization of protein sequences. I have been downloading sequences from UniProt and ran weka, but accuracy is constantly appearing around 30-40%. Is there any way to improve it by making dataset accordingly?

Similar questions and discussions