seeking advice on how to configure Phaser-MR (full-featured) to use a previous partial solution
Hello. I have a 3-domain protein (let's say it's organized from N to C as A-B-C) whose structure I am working to determine now. Although several single- (A alone, B alone), double- (A+B and B+C) and 3-domain (A+B+C) structures of this protein (from a different organism whose primary sequence is, overall, ~85% identical to the one on which I'm working) have already been determined and deposited in the PDB, I would ideally like to use a minimal probe (i.e. as few domains as possible) in order avoid model bias that could be imposed on the relative orientation of domains A, B, and C. So far, I have phase solutions (TFZ scores for all in the high-50s and 60s) using B-only, A+B, and B+C pdb files as probes. A-only structures have not given a phase solution (TFZ is ~7 and autobuild gives a solution that looks unbelievable in Coot), perhaps because A is relatively small in size. Here is my problem: I suspect that A and B sit relative to one another in an interesting way in this particular structure, so I'd like to use Phaser to search for a solution given the fixed partial solution from a B-only search, together with an A-only structure as a separate probe. So far no MR solutions have been found when I configure the run with an A-only pdb file and a B-only-derived Phaser result that is input as a fixed partial solution. I'm writing to ask for advice on how to configure the search. I think that the problem is my configuring the search, and not that a solution cannot be found from these coordinates. This is the reason I think that my problem is in configuring the search; please correct me if I'm wrong: Using a Phaser result derived from using an A+B structure as probe, I cut up the resulting pdb file into 2 pdb files - one containing an A' and another containing a B', where A' is the A domain of the Phaser result, and B' is the B-only part. Then I configured a Phaser result with B' as an ensemble that is a fixed partial solution, and A' as a separate ensemble/probe. I searched for 4 of each since 4 chains is the most likely composition in the ASU. I expected to get a result similar to what Phaser calculated when I input the A+B structure as an intact probe. Instead of getting a solution where 4 A+B chains were placed as they were in the first Phaser result, which gave one solution with a TFZ score of 65, I got 13 solutions, the top scoring a TFZ of 7.5, and each solution having a total of 9 chains - 4 As and 5 Bs. Is anyone willing to help me troubleshoot the above before I throw up my hands and settle on either manually building two domains onto the single-domain solution, or one domain onto the 2-domain solution, keeping in mind the potential for model bias? I am keen on using the B-only solution as a fixed partial solution to place domain A because all the B-only structures provide coordinate information for a portion of domain B that is unresolved in any of the multi-domain structures that contain domain B. I am using the Phenix gui, version 1.12rc1-2807. Regards, Emily. -- "Study as if you were going to live forever; live as if you were going to die tomorrow." - Maria Mitchell "I'm not afraid of storms for I'm learning to sail my ship." - Louisa May Alcott
Hi, I think that the Phaser tutorial is your friend here. There is a nice example of how to run two successive searches in one run to solve a n-domain structure (we actually used it recently to solve a structure of a protein-DNA complex). I think that running your B+C search followed by an A search would solve your problem. If you haven't tried it yet it's worth trying. Good luck, Boaz Boaz Shaanan, Ph.D. Dept. of Life Sciences Ben-Gurion University of the Negev Beer-Sheva 84105 Israel E-mail: [email protected] Phone: 972-8-647-2220 Skype: boaz.shaanan Fax: 972-8-647-2992 or 972-8-646-1710 ________________________________________ From: [email protected] [[email protected]] on behalf of Emilia C. Arturo (Emily) [[email protected]] Sent: Tuesday, August 08, 2017 7:00 PM To: PHENIX user mailing list Subject: [phenixbb] seeking advice on how to configure Phaser-MR (full-featured) to use a previous partial solution Hello. I have a 3-domain protein (let's say it's organized from N to C as A-B-C) whose structure I am working to determine now. Although several single- (A alone, B alone), double- (A+B and B+C) and 3-domain (A+B+C) structures of this protein (from a different organism whose primary sequence is, overall, ~85% identical to the one on which I'm working) have already been determined and deposited in the PDB, I would ideally like to use a minimal probe (i.e. as few domains as possible) in order avoid model bias that could be imposed on the relative orientation of domains A, B, and C. So far, I have phase solutions (TFZ scores for all in the high-50s and 60s) using B-only, A+B, and B+C pdb files as probes. A-only structures have not given a phase solution (TFZ is ~7 and autobuild gives a solution that looks unbelievable in Coot), perhaps because A is relatively small in size. Here is my problem: I suspect that A and B sit relative to one another in an interesting way in this particular structure, so I'd like to use Phaser to search for a solution given the fixed partial solution from a B-only search, together with an A-only structure as a separate probe. So far no MR solutions have been found when I configure the run with an A-only pdb file and a B-only-derived Phaser result that is input as a fixed partial solution. I'm writing to ask for advice on how to configure the search. I think that the problem is my configuring the search, and not that a solution cannot be found from these coordinates. This is the reason I think that my problem is in configuring the search; please correct me if I'm wrong: Using a Phaser result derived from using an A+B structure as probe, I cut up the resulting pdb file into 2 pdb files - one containing an A' and another containing a B', where A' is the A domain of the Phaser result, and B' is the B-only part. Then I configured a Phaser result with B' as an ensemble that is a fixed partial solution, and A' as a separate ensemble/probe. I searched for 4 of each since 4 chains is the most likely composition in the ASU. I expected to get a result similar to what Phaser calculated when I input the A+B structure as an intact probe. Instead of getting a solution where 4 A+B chains were placed as they were in the first Phaser result, which gave one solution with a TFZ score of 65, I got 13 solutions, the top scoring a TFZ of 7.5, and each solution having a total of 9 chains - 4 As and 5 Bs. Is anyone willing to help me troubleshoot the above before I throw up my hands and settle on either manually building two domains onto the single-domain solution, or one domain onto the 2-domain solution, keeping in mind the potential for model bias? I am keen on using the B-only solution as a fixed partial solution to place domain A because all the B-only structures provide coordinate information for a portion of domain B that is unresolved in any of the multi-domain structures that contain domain B. I am using the Phenix gui, version 1.12rc1-2807. Regards, Emily. -- "Study as if you were going to live forever; live as if you were going to die tomorrow." - Maria Mitchell "I'm not afraid of storms for I'm learning to sail my ship." - Louisa May Alcott _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]
participants (2)
-
Boaz Shaanan -
Emilia C. Arturo (Emily)