User:Wayne Decatur/1ig8 to 3b8a (hexokinase) morph methods

From Proteopedia

Revision as of 03:18, 9 January 2012

MORPH

• First I changed chain X of 3b8a to chain A to match 1ig8 using PDB Goodies for the protein. I knew PDB Goodies ignored and left out the heteroatoms so I changed the few lines of the glucose ligand to chain a by hand. The heterotoms will not be used by the Yale Morph server but I need them in the original file in order to be able to use fit or compare to add them in later.
• Submitted 1gi8 and the altered 3b8a to the Yale Morph beta server
• TURNS OUT THIS NEXT ISSUE WAS NOT INVOLVED IN PROBLEM SO CAN SKIP TO TWO STEPS BELOW WHERE I WAS ABLE TO GET RESULT. ---When I received nothing back after several minutes I looked at the end of the sequences to see if they matched identially and noticed they didn't have the same C-terminus or even N-terminus. In fact what I thought were the same protein, according to FASTA sequence obtained for each PDB file from PDBsum, differ quite a bit:

1ig8 1 ---DVPKELMQQIENFEKIFTVPTETLQAVTKHFISELEKGLSKKGGNIP 47 |||||||.:|...|.:|||.:|||:.|.||||.||.|||:||||||| 3b8a 1 SMADVPKELMDEIHQLEDMFTVDSETLRKVVKHFIDELNKGLTKKGGNIP 50 1ig8 48 MIPGWVMDFPTGKESGDFLAIDLGGTNLRVVLVKLGGDRTFDTTQSKYRL 97 |||||||:||||||||::|||||||||||||||||.|:.|||||||||:| 3b8a 51 MIPGWVMEFPTGKESGNYLAIDLGGTNLRVVLVKLSGNHTFDTTQSKYKL 100 1ig8 98 PDAMRTTQNPDELWEFIADSLKAFIDEQFPQGISEPIPLGFTFSFPASQN 147 |..||||::.:|||.|||||||.|:.||.... ::.:|||||||:||||| 3b8a 101 PHDMRTTKHQEELWSFIADSLKDFMVEQELLN-TDTLPLGFTFSYPASQN 149 1ig8 148 KINEGILQRWTKGFDIPNIENHDVVPMLQKQITKRNIPIEVVALINDTTG 197 ||||||||||||||||||:|.|||||:||.:|:||.:|||:|||||||.| 3b8a 150 KINEGILQRWTKGFDIPNVEGHDVVPLLQNEISKRELPIEIVALINDTVG 199 1ig8 198 TLVASYYTDPETKMGVIFGTGVNGAYYDVCSDIEKLQGKLSDDIPPSAPM 247 ||:||||||||||||||||||||||:|||.||||||:|||:||||.::|| 3b8a 200 TLIASYYTDPETKMGVIFGTGVNGAFYDVVSDIEKLEGKLADDIPSNSPM 249 1ig8 248 AINCEYGSFDNEHVVLPRTKYDITIDEESPRPGQQTFEKMSSGYYLGEIL 297 |||||||||||||:||||||||:.:||:|||||||.||||:|||||||:| 3b8a 250 AINCEYGSFDNEHLVLPRTKYDVAVDEQSPRPGQQAFEKMTSGYYLGELL 299 1ig8 298 RLALMDMYKQGFIFKNQDLSKFDKPFVMDTSYPARIEEDPFENLEDTDDL 347 ||.|:::.::|.:.|:|||||..:|::||||||||||:|||||||||||: 3b8a 300 RLVLLELNEKGLMLKDQDLSKLKQPYIMDTSYPARIEDDPFENLEDTDDI 349 1ig8 348 FQNEFGINTTVQERKLIRRLSELIGARAARLSVCGIAAICQKRGYKTGHI 397 ||.:||:.||:.||||||||.||||.|||||:|||||||||||||||||| 3b8a 350 FQKDFGVKTTLPERKLIRRLCELIGTRAARLAVCGIAAICQKRGYKTGHI 399 1ig8 398 AADGSVYNRYPGFKEKAANALKDIYGWTQTSLDDYPIKIVPAEDGSGAGA 447 ||||||||:||||||.||..|:||||||..:..| ||.|||||||||||| 3b8a 400 AADGSVYNKYPGFKEAAAKGLRDIYGWTGDASKD-PITIVPAEDGSGAGA 448 1ig8 448 AVIAALAQKRIAEGKSVGIIGA 469 ||||||::||||||||:||||| 3b8a 449 AVIAALSEKRIAEGKSLGIIGA 470

Turns out they are not the same protein; one is yeast hexokinase PI (3b8a) and the other is hexokinase PII (1ig8).

According to help page for the Yale Morph server this task was submitted, this shouldn't matter because it will do a ClustalW alignment. Frustratingly, though it does seem to matter because they always seem to need to be identical to me for anything to work at the Yale Morph server. And unfortunately when it fails, it doesn't tell you why it failed or even that it failed. A lot of stuff they talk about never got implemented and maybe the alignment was part of it?<--NOPE, IT WAS A SERVER GLITCH. SEE NEXT TO SECTIONS.

• THIS WHOLE AVENUE WAS UNNECESSARY BUT IT HELPED LEAD TO ISSUE.---To get so they were the same two proteins, I submitted the FASTA sequence of 3b8a (without amino acids 15-17) (that i got from PDBsum) to SwissModel and told it to homology model it to 1ig8. Got that result and called 3b8amodeledto1ig8.pdb
  • Submitted 3b8amodeledto1ig8.pdb and the 3b8achainA file to every version of the Yale Morph server I could find. However, I got nothing back for any attempt and when I looked at the new file I realized it had renumbered to start at 1, and so to get equal chains, I submitted the FASTA sequence of 3b8a (without amino acids 15-17) (that i got from PDBsum) to SwissModel and told it to model it to 3b8aASchainA.pdb. This should renumber like it did when I submitted it to model against 1ig8. By doing the chain A version, I hope it helps insure I stay dealing with that chain only.
  • Submitted the set of files on every version and it still failed to give me any completed job message.
  • Then I tried some things that I knew worked in the past and they failed to elicit e-mails and so I realized server had problems and contacted them. They fixed something and I still wasn't receiving e-mail replies of completion. Decided to use an old job complete message I fortunately had and changed the URL to include the new submission confirmation IDs and I was able to get the test files.
  • I informed the server people that maybe e-mails weren't being sent beyond Yale and they said that seemed to be problem and would work on it but I had a work around for now of updating old URL with new submissions.

• So I resubmitted 1gi8 and the altered 3b8a (altered to chain A) to the Yale Morph beta server and noted the confirmation number. Later tried th confirmation number in the URL (http://www.molmovdb.org/cgi-bin/morph.cgi?ID=b049347-636) and result was there.
  • Downloaded result.
  • Edited model numbers to go from 1-12 instead of 0 to 11 since the 0 model number causes issues when I try to call it in Jmol. --> file = 1ig8TO3b8a.pdb.