update alphabet

sequence_models/constants.py

@@ -27,6 +27,7 @@ PAD = GAP
 MSA_PAD = '!'
 MASK = '#'  # Useful for masked language model training
 START = '@'
+SEP = '/'
 
 SPECIALS = STOP + GAP + MASK + START
 PROTEIN_ALPHABET = ALL_AAS + SPECIALS
@@ -34,6 +35,7 @@ MSA_AAS = ALL_AAS + GAP
 MSA_ALPHABET = ALL_AAS + GAP + STOP + MASK + START + MSA_PAD
 RNA_ALPHABET = IUPAC_AMB_RNA + SPECIALS
 ENHANCER_ALPHABET = DNA + 'N' + GAP + MSA_PAD + MASK + STOP + START
+MSA_ALPHABET_PLUS = MSA_ALPHABET + SEP
 
 trR_ALPHABET = "ARNDCQEGHILKMFPSTWYV-"