GENN+ASAquick: A general neural network and a quick ASA predictor (GENN PAPER) (ASAquick PAPER)
NOTE: For training purposes the ASA was scaled in ASAquick as
RASA' = 2 * ( ASA / ASAm(R) - 0.5), with ASAm(R) the maximum ASA value for residue type R.
Since the standard definition of RASA is
RASA = ASA / ASAm(R),
the relationship between them is:
RASA = 0.5 + RASA' / 2,
RASA' = 2 * (RASA - 0.5).