Ran LOMETS with query
>SPU_000100.1 Peptide Sequence
AIGTVGKVVEIDGSDIRVKFKDNQNFIFNSVNLTPERIRGPKEEALNLEKKLRGVTRILKGNEWSVVTAADTGDIKAVERILDKHPEWGHCNVNVVDSEMGVCPLAFAISGDNFHIVDVLVEHGADVNLPTLSGNTSLQVALNVRYSDTPSQNMKEEPYMAEMKRKWGHLGIETNSDALLAFLVSHGGDLNSFNGDNETIKSLLQHDPILQPLLEIQSQRGRDIARQQSTANRPLRSLTKKTNDTATEVNAVSSLLQQSTARPSSKKLSSGNSSEKLATGSGNQRKSSGELHHLQAELQRKDEELETLRSRVKTLEHLVKKRISFACGHDVIRSIAELIPECPNCSEPISSVVDLQ
and it had a bunch of hits with typical PDB IDs, plus one on "4o1oA1". PDB entry 4O1O has a chain A with residues 22 to 321, a large gap, and residues 333-729. 4o1oA1 appears to correspond to residues 22-321 of 4010 chain A (and I didn't inspect the rest of chain A more carefully).
So it looks like the PDB database LOMETS used at some point broke chain A at that big gap and numbered the pieces. Looked in the PDB database and found that it actually contains:
127384 4o1oA1.pdb
45489 4o1oA2.pdb
63749 4o1oA3.pdb
57754 4o1oA4.pdb
294364 4o1oA.pdb
where 4o1oA contains at least 4o1oA1, and presumably all of the others, but it is a little hard to tell because
the broken up pieces have all been renumbered from 1.
What are the rules for when this happens? Are there any other oddities in that database, that would result in an ID other than the usual 4 digit PDB entry followed by the chain letter? How did LOMETS decide to use 4o1oA1 and not 4o1oA, when the blast output was in terms of a gi?
Thanks.