Breakdown(get the substructure of) the glycan that only has topology information, no linkage info

[bobaoai]

Hi, I am using this code to get all possible substructures in "_glycan". I cannot get the string of the substructure.

_glycan=glycoct.loads("""RES
1b:x-dglc-HEX-1:5
2b:x-lgal-HEX-1:5|6:d
3b:x-dglc-HEX-1:5
4b:x-dman-HEX-1:5
5b:x-dman-HEX-1:5
6b:x-dglc-HEX-1:5
7b:x-dgal-HEX-1:5
8b:a-dgro-dgal-NON-2:6|1:a|2:keto|3:d
9s:n-acetyl
10s:n-acetyl
11b:x-dman-HEX-1:5
12b:x-dglc-HEX-1:5
13b:x-dgal-HEX-1:5
14s:n-acetyl
15b:x-dglc-HEX-1:5
16b:x-dgal-HEX-1:5
17s:n-acetyl
18s:n-acetyl
19s:n-acetyl
LIN
1:1o(-1+1)2d
2:1o(-1+1)3d
3:3o(-1+1)4d
4:4o(-1+1)5d
5:5o(-1+1)6d
6:6o(-1+1)7d
7:7o(3|6+2)8d
8:8d(5+1)9n
9:6d(2+1)10n
10:4o(-1+1)11d
11:11o(-1+1)12d
12:12o(-1+1)13d
13:12d(2+1)14n
14:11o(-1+1)15d
15:15o(-1+1)16d
16:15d(2+1)17n
17:3d(2+1)18n
18:1d(2+1)19n
”“”)

_frag_motif_list = {}
    for i in _glycan.fragments(max_cleavages=len(_glycan)):
        _frag_gly = fragment_to_substructure(i, _glycan)
        if not len(_frag_gly) in _frag_motif_list.keys():
            _frag_motif_list[len(_frag_gly)] = [glycoct.loads(str(_frag_gly))]
        else:
            _frag_motif_list[len(_frag_gly)].append(glycoct.loads(str(_frag_gly)))

I tried to print the node that generated the error by inserting some print() in source code: The error returns: RES 1b:x-dglc-HEX-1:5 2s:n-acetyl LIN 1:1d(2+1)2n 2 RES 1b:x-dglc-HEX-1:5 2s:n-acetyl LIN 1:1d(2+1)2n 2 tag 0 (4)o(-1+1)d(5)[x] (5)o(-1+1)d(6)[x] None

TypeError Traceback (most recent call last)

in () 25 # print('finished getmotif') 26 ---> 27 get_motif(glycoct.loads(a_3350),1) 28 # str() in get_motif(glycoct_obj, idex) 15 _frag_motif_list[len(_frag_gly)] = [glycoct.loads(str(_frag_gly))] 16 else: ---> 17 _frag_motif_list[len(_frag_gly)].append(glycoct.loads(str(_frag_gly))) 18 # except: 19 # plot(_frag_gly) ~/anaconda3/lib/python3.5/site-packages/glypy/structure/glycan.py in serialize(self, name) 673 674 def serialize(self, name='glycoct'): --> 675 return self._serializers[name](self) 676 677 __repr__ = serialize ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in dumps(structure, full) 1225 1226 def dumps(structure, full=True): -> 1227 return GlycoCTWriter(structure, None, full=full).dump() 1228 1229 ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in dump(self) 960 961 def dump(self): --> 962 buffer = self.handle_glycan() 963 if self.nobuffer: 964 value = buffer.getvalue() ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in handle_glycan(self) 1182 visited.add(link.child.id) 1183 if link.child.node_type is Monosaccharide.node_type: -> 1184 line = self.handle_monosaccharide(link.child) 1185 else: 1186 line = self.handle_substituent(link.child) ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in handle_monosaccharide(self, monosaccharide) 1144 link_collection.extend([cl for p, cl in monosaccharide.children(links=True)]) 1145 -> 1146 links = self.ordering_context.sort_links(link_collection) 1147 self.link_queue.extend(links) 1148 return residue_str ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in sort_links(self, links) 1080 1081 def sort_links(self, links): -> 1082 return sorted(links, key=cmp_to_key(self.compare_link_ordering)) 1083 1084 def sort_residues(self, residues): ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in compare_link_ordering(self, link_a, link_b) 1075 print(child_a) 1076 print(child_b) -> 1077 ordered = self.compare_residue_ordering(child_a, child_b) 1078 1079 return ordered ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in compare_residue_ordering(self, res_a, res_b) 1037 return diff_n_branches_from 1038 -> 1039 subtree_a = GlycoCTWriter(Glycan.subtree_from(self.structure, res_a)).dump() 1040 subtree_b = GlycoCTWriter(Glycan.subtree_from(self.structure, res_b)).dump() 1041 return subtree_a < subtree_b ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in dump(self) 960 961 def dump(self): --> 962 buffer = self.handle_glycan() 963 if self.nobuffer: 964 value = buffer.getvalue() ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in handle_glycan(self) 1168 visited = set() 1169 if self.structure.root.node_type is Monosaccharide.node_type: -> 1170 res_str = self.handle_monosaccharide(self.structure.root) 1171 self.buffer.write(res_str + "\n") 1172 else: ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in handle_monosaccharide(self, monosaccharide) 1144 link_collection.extend([cl for p, cl in monosaccharide.children(links=True)]) 1145 -> 1146 links = self.ordering_context.sort_links(link_collection) 1147 self.link_queue.extend(links) 1148 return residue_str ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in sort_links(self, links) 1080 1081 def sort_links(self, links): -> 1082 return sorted(links, key=cmp_to_key(self.compare_link_ordering)) 1083 1084 def sort_residues(self, residues): ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in compare_link_ordering(self, link_a, link_b) 1075 print(child_a) 1076 print(child_b) -> 1077 ordered = self.compare_residue_ordering(child_a, child_b) 1078 1079 return ordered ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in compare_residue_ordering(self, res_a, res_b) 1025 print(link) 1026 if link.is_parent(res_a): -> 1027 branch_label = self.get_branch_from_link_label(link) 1028 n_branches_from_a = max(n_branches_from_a, self.branch_to_terminal_count[branch_label]) 1029 ~/anaconda3/lib/python3.5/site-packages/glypy/io/glycoct.py in get_branch_from_link_label(self, link) 984 def get_branch_from_link_label(self, link): 985 print(link.label) --> 986 return link.label[0] 987 988 def build_branch_to_terminal_count(self): TypeError: 'NoneType' object is not subscriptable

[mobiusklein]

This error is caused by GlycoCT serialization code assuming that the Glycan object was indexed already. I've added a check for this and it will index the glycan automatically.

There is a method to do this subtree enumeration directly without needing to resort to fragment_to_substructure, Glycan.substructures. It uses the same machinery as Glycan.fragments, but skips the extra work that fragments does to account for the bond cleavage losses from particular types of dissociation in a mass spectrometer.