Table 1.

Average values of properties calculated with PepFun for the datasets of peptides associated to the MHC class II protein and the Granzyme B protease. We report the average net charge based on pka values of each amino acid at pH 7, the molecular weight calculated in g/mol using the SMILES representation of the peptide, the Crippen estimation of the octanol/water partition coefficient (logP) available in the RDKit [24], the average hydrophobicity from the Eisenberg scale [23], the isoelectric point and aromaticity obtained from the ProtParam package [9], the instability index from ProtParam as an estimate of the stability of the peptide in a test tube [9], the number of hydrogen bond acceptors and donors calculated using the SMILES representation, and the number of failed solubility and synthesis empirical rules [26]. We note that the higher the number of rules violated, the lower the probability to be solubilized or synthesized experimentally.

Property	Average Values MHC Set	Average Values Protease Set
Net charge	0.457	−1.872
Molecular weight (g/mol)	1696.111	884.133
LogP	−5.810	−4.235
Hydrophobicity	1.735	0.002
Aromaticity	0.099	0.043
Instability index	33.883	37.095
Isoelectric point	7.290	4.375
Number hydrogen donors	24.183	13.337
Number hydrogen acceptors	23.670	13.578
Number of solubility rules failed	2	1.5
Number of synthesis rules failed	4.5	0.5