November 2004 - modeller_usage

Re: RE: [modeller_usage] wrong topology for the ligand?
by Cvetan Stojkoski 17 Nov '04

17 Nov '04

Quoting Youbin Tu <ytu(a)mix.wvu.edu>: > Hi Alicia, Ben: > Thanks for your reply, which help me a lot. > But for my specific situation, the ligand I am trying to use is > phenacetin which did not exist in the template pdb file. Or to say, we > created it so as to make it docked in the active site by using some NMR > restraints. We expected to see some interactions between the ligand and > the active site which will make the active site show a differnt look > from what it is when no ligand is bound. You have used NMR perturbation data to dock a ligand into the active site? You could simulate a ligand-induced conformational change in the receptor just by doing some MD on your complex. I recommend NAMD. However, AMBER will probably work better for you since your compound has an aromatic ring. > Now I was really confused, if the ligand can not have correct > connectivity, how can we expect to see the interactions between the > ligand and the enzyme? I mean how the follwong refinement steps which > is involved with minimization and molecular dynamics correctly dealt > with the ligand and the interaction between the ligand and the enzyme? MD packages rely on complete topology/parameter sets. You can use XPLO2D to predict a lot of this information. However, be aware that the values are completely derived from the pdb file and will need to be edited. Charges will also need to be added. > Again, why the ligand does not show correctly, you know , we build > the phenacetin in Builder module in InsightII and output it as a RTF > file, then append it to the top_heav.lib file. The molecule looks very > reasonable in InsightII, but after calculation by MODELLER , the ligand > looked weird. So, any other program can view RTF file so as to have a > check the connectivity of the ligand is reasonably good or not? Or if > the parameters are incomplete, how to make it complete? Although we have InsightII, I have never built an RTF file with it. I'm guessing it works similar to XPLO2D. Therefore you will most certainly have to edit angle and charge terms. However, you'd think that bonding would be correct. > Another question is whether does the RTF file provide enough > connectivity information for the ligand? Or is it because of the > special restraints we applied in MODELLER which make the connectivity > changed? You know,we even tried to use the ligand without any > restraints, the result still looked weird? To check if your RTF file provides correct connectivity, pass it through X-PLOR along with your pdb file to produce a psf. You can then view your psf file in VMD to make sure your connectivities are correct. Otherwise just go through the topology manualy. > Basically, we just want to build a new ligand with some experimental > restraints in the active site of a enzyme. If the ligand looked werid, > we have to doubt whether the interaction between the ligand and the > enzyme is reliable or not? So, we have to know which step create this > problem? If you already have the structure of the protein you are docking to then I wouldn't recommend using MODELLER to dock your ligand and observe ligand-induced confomrational changes. Do MD. > 1. building the ligand (RFT file)? We did not give enough connectivity > information? > 2. running MODELLER? Since there is no restraints to confine the bond > or angles for the ligand or the restraints are conflicting with each > other which make the ligand broken? > 3. outputting file from MODELLER? Assumed that all the above is good, > the only problem is that MODDLER only give the cooridinates of the > atoms instead of the relationship ( connectivity) between atoms? > Hope I have well explained my problems? And I also attached my RTF > file for phenacetin. > Finally, thanks for your guys' kind help. Anyone's reply in this > topic will be highly appreciated. > Hope for the best. > > youbin > Sorry I didn't go through your RTF file but its laborious work even with the smallest of compounds. Best create the psf file and then visualize it in VMD.

1 0

Re: RE: [modeller_usage] wrong topology for the ligand?
by Youbin Tu 17 Nov '04

17 Nov '04

Hi Alicia, Ben: Thanks for your reply, which help me a lot. But for my specific situation, the ligand I am trying to use is phenacetin which did not exist in the template pdb file. Or to say, we created it so as to make it docked in the active site by using some NMR restraints. We expected to see some interactions between the ligand and the active site which will make the active site show a differnt look from what it is when no ligand is bound. Now I was really confused, if the ligand can not have correct connectivity, how can we expect to see the interactions between the ligand and the enzyme? I mean how the follwong refinement steps which is involved with minimization and molecular dynamics correctly dealt with the ligand and the interaction between the ligand and the enzyme? Again, why the ligand does not show correctly, you know , we build the phenacetin in Builder module in InsightII and output it as a RTF file, then append it to the top_heav.lib file. The molecule looks very reasonable in InsightII, but after calculation by MODELLER , the ligand looked weird. So, any other program can view RTF file so as to have a check the connectivity of the ligand is reasonably good or not? Or if the parameters are incomplete, how to make it complete? Another question is whether does the RTF file provide enough connectivity information for the ligand? Or is it because of the special restraints we applied in MODELLER which make the connectivity changed? You know,we even tried to use the ligand without any restraints, the result still looked weird? Basically, we just want to build a new ligand with some experimental restraints in the active site of a enzyme. If the ligand looked werid, we have to doubt whether the interaction between the ligand and the enzyme is reliable or not? So, we have to know which step create this problem? 1. building the ligand (RFT file)? We did not give enough connectivity information? 2. running MODELLER? Since there is no restraints to confine the bond or angles for the ligand or the restraints are conflicting with each other which make the ligand broken? 3. outputting file from MODELLER? Assumed that all the above is good, the only problem is that MODDLER only give the cooridinates of the atoms instead of the relationship ( connectivity) between atoms? Hope I have well explained my problems? And I also attached my RTF file for phenacetin. Finally, thanks for your guys' kind help. Anyone's reply in this topic will be highly appreciated. Hope for the best. youbin Higueruelo Alicia (CAM) wrote: >Hi Youbin, > >In my experience,(please someone correct me if I'm wrong) the ligands in >modeller don't look great, my guess is that CHARMm deals really well with >aminoacids but struggle a bit with non-aminoacid atom types. (or a least the >version I'm using). > >But they look reasonable. If the ligand look really really weird. try to >compare the "PDB atom names" in your template ( the pdb file you are using >to built your model) and the "PDB atom names" in your top_heavy.lib. Hope >the following lines are a good example. > >in top_heav.lib in RESI LIG "ATOM O1 OE -0.10000" >in template.pdb "ATOM 4000 O1 LIG D 500 30.722 47.505 16.832" > >To have the same connection and geometry in the model-ligand than in the >ligand-template, these PDB names should be the same "O1", I think this is >what is telling Modeller where to put the atom (assigning the xyz coords >from atom called O1 from the LIG in the PDB), and "OE" is the CHARMm atom >type to use the potentials . > >In my output files, my model does not have connection data in the pdb, >therefore the way you see you models depends which software you are using to >visualize your pdb, again, lots of them can not guess that a phenyl ring >(not in a aminoacid) has alternate double bounds or aromatic bonds, and >display them in funny ways. There are spl macros in sybyl for example, which >will fix some of these problems, based on the geometry of the ligands. but >I'm not aware of a 3D viewer which display chemically correctly all ligands >in pdb format. > >Hope this is useful, > >any comments on that more than welcome, > >cheers, > >alicia > > >-----Original Message----- >From: Youbin Tu [mailto:ytu@mix.wvu.edu] >Sent: 16 November 2004 07:08 >To: modeller_usage(a)salilab.org >Subject: [modeller_usage] wrong topology for the ligand? > > >Hi all: > > I am using MODELLER7V7 to build a ligand-enzyme complex, I created >the ligand and modified the library files. Everything worked well >except the ligand in the final result seems too weird, for example the >benzene is not planar and there are many bonds between 2 atoms which >should only have one bond theoritically. > I thought maybe it is my problem when building the ligand. But >when I tried to use GDP which already existed in the library file. The >final >results still have the same problem. > Anyone knows how to deal with this kind of problem? Does it mean >that we have to modify the par.lib or make restraints for bonds of the >ligands? If so, can you tell me how to do that,any softwares are >available for that purpose? > Your reply is highly appreciated. > > >youbin > > > > > >_______________________________________________ >modeller_usage mailing list >modeller_usage(a)salilab.org >http://salilab.org/mailman/listinfo/modeller_usage > > >The information contained in this email is intended for the >personal and confidential use of the addressee only. It may >also be privileged information. If you are not the intended >recipient then you are hereby notified that you have received >this document in error and that any review, distribution or >copying of this document is strictly prohibited. If you have >received this communication in error, please notify Celltech >immediately on: >+44 (0)1753 534655, or email 'is(a)celltech.co.uk' >Celltech Group plc >208 Bath Road, Slough, SL1 3WE, Berkshire, UK >Registered Office as above. > > ---------------------- Youbin Tu CCMM Lab,BPS Schoof of Pharmacy,WVU

2 1

Problems with Modeller 7v7 in windows XP
by Alfonso Mendez 16 Nov '04

16 Nov '04

Hi friends: After the installation of modeller in my PC I tried to follow a similar example like those are illustrated in the WEB page. I prepared a file liki this: XUFF11.ali >P1;XUFF11 sequence:XUFF11:::::::0.00: 0.00 MVDFYYLPGSSPCRSVIMTAKAVGVELNKKLLNLQAGEHLKPEFLKINPQHTIPTLVDNGFALWESRAIQVYLVEKYGKTDSLY PKCPKKRAVINQRLYFDMGTLYQSFANYYYPQVFAKAPADPEAFKKIEAAFEFLNTFLEGQDYAAGDSLTVADIALVATVSTFE VAKFEISKYANVNRWYENAKKVTPGWEENWAGCLEFKKYFE* and a script file like this: SET SEARCH_RANDOMIZATIONS = 100 SET FILE = 'XUFF11.ali' READ_SEQUENCE_DB SEQUENCE_SEARCH ALIGN_CODES = 'XUF11', DATA_FILE = ON The program worked and after an hour the process finished. The log file showed this information in the last lines: ddgrp__223E> Array too small. Increase MAXSEQ current maximum, current need: 80 115 addgrp__223E> Array too small. Increase MAXSEQ current maximum, current need: 80 87 addgrp__223E> Array too small. Increase MAXSEQ current maximum, current need: 80 83 addgrp__223E> Array too small. Increase MAXSEQ current maximum, current need: 80 84 Then I tried to run the following script: READ_MODEL FILE = '1jlv.pdb' SEQUENCE_TO_ALI ALIGN_CODES = '1jlv' READ_ALIGNMENT FILE = 'XUFF11.ali', ALIGN_CODES = 'XUFF11', ADD_SEQUENCE = ON ALIGN2D WRITE_ALIGNMENT FILE='XUFF11-1jlv.ali', ALIGNMENT_FORMAT = 'PIR' having the 1jlv.pdb file in the same directory that the script files and then i received the next message in the console: C:\bioinfo\Modeller7v7\examples\commands>mod7v7 align2d.top forrtl: severe (170): Program Exception - stack overflow Image PC Routine Line Source mod7v7.exe 006045E6 Unknown Unknown Unknown mod7v7.exe 004D1CAF Unknown Unknown Unknown mod7v7.exe 004066D8 Unknown Unknown Unknown mod7v7.exe 0040F5BB Unknown Unknown Unknown mod7v7.exe 006A564B Unknown Unknown Unknown mod7v7.exe 006EDB39 Unknown Unknown Unknown mod7v7.exe 006DC824 Unknown Unknown Unknown kernel32.dll 7C816D4F Unknown Unknown Unknown and the content of the log file was: MODELLER 7v7, Sep 12, 2004 09:15pm PROTEIN STRUCTURE MODELLING BY SATISFACTION OF SPATIAL RESTRAINTS ..... Kind, OS, HostName, Kernel, Processor: 4, WinXP build 2600 Service Pack 2, HPPAVILION, uni, x86 Family 15 Model 2 Stepping 9 Date and time of compilation : 09/13/2004 17:17:22 Job starting time (YY/MM/DD HH:MM:SS): 2004/11/16 22:56:43.406 My queestion is What happened? Alfonso Mendez Tenorio Escuela Nacional de Ciencias Biológicas Mexico city. --------------------------------- Do You Yahoo!? Yahoo! Net: La mejor conexión a internet y 2GB en tu buzón de Correo Yahoo! por $100 al mes.

1 0

wrong topology for the ligand?
by Youbin Tu 16 Nov '04

16 Nov '04

Hi all: I am using MODELLER7V7 to build a ligand-enzyme complex, I created the ligand and modified the library files. Everything worked well except the ligand in the final result seems too weird, for example the benzene is not planar and there are many bonds between 2 atoms which should only have one bond theoritically. I thought maybe it is my problem when building the ligand. But when I tried to use GDP which already existed in the library file. The final results still have the same problem. Anyone knows how to deal with this kind of problem? Does it mean that we have to modify the par.lib or make restraints for bonds of the ligands? If so, can you tell me how to do that,any softwares are available for that purpose? Your reply is highly appreciated. youbin

3 2

RE: [modeller_usage] wrong topology for the ligand?
by Higueruelo Alicia (CAM) 16 Nov '04

16 Nov '04

Hi Youbin, In my experience,(please someone correct me if I'm wrong) the ligands in modeller don't look great, my guess is that CHARMm deals really well with aminoacids but struggle a bit with non-aminoacid atom types. (or a least the version I'm using). But they look reasonable. If the ligand look really really weird. try to compare the "PDB atom names" in your template ( the pdb file you are using to built your model) and the "PDB atom names" in your top_heavy.lib. Hope the following lines are a good example. in top_heav.lib in RESI LIG "ATOM O1 OE -0.10000" in template.pdb "ATOM 4000 O1 LIG D 500 30.722 47.505 16.832" To have the same connection and geometry in the model-ligand than in the ligand-template, these PDB names should be the same "O1", I think this is what is telling Modeller where to put the atom (assigning the xyz coords from atom called O1 from the LIG in the PDB), and "OE" is the CHARMm atom type to use the potentials . In my output files, my model does not have connection data in the pdb, therefore the way you see you models depends which software you are using to visualize your pdb, again, lots of them can not guess that a phenyl ring (not in a aminoacid) has alternate double bounds or aromatic bonds, and display them in funny ways. There are spl macros in sybyl for example, which will fix some of these problems, based on the geometry of the ligands. but I'm not aware of a 3D viewer which display chemically correctly all ligands in pdb format. Hope this is useful, any comments on that more than welcome, cheers, alicia -----Original Message----- From: Youbin Tu [mailto:ytu@mix.wvu.edu] Sent: 16 November 2004 07:08 To: modeller_usage(a)salilab.org Subject: [modeller_usage] wrong topology for the ligand? Hi all: I am using MODELLER7V7 to build a ligand-enzyme complex, I created the ligand and modified the library files. Everything worked well except the ligand in the final result seems too weird, for example the benzene is not planar and there are many bonds between 2 atoms which should only have one bond theoritically. I thought maybe it is my problem when building the ligand. But when I tried to use GDP which already existed in the library file. The final results still have the same problem. Anyone knows how to deal with this kind of problem? Does it mean that we have to modify the par.lib or make restraints for bonds of the ligands? If so, can you tell me how to do that,any softwares are available for that purpose? Your reply is highly appreciated. youbin _______________________________________________ modeller_usage mailing list modeller_usage(a)salilab.org http://salilab.org/mailman/listinfo/modeller_usage The information contained in this email is intended for the personal and confidential use of the addressee only. It may also be privileged information. If you are not the intended recipient then you are hereby notified that you have received this document in error and that any review, distribution or copying of this document is strictly prohibited. If you have received this communication in error, please notify Celltech immediately on: +44 (0)1753 534655, or email 'is(a)celltech.co.uk' Celltech Group plc 208 Bath Road, Slough, SL1 3WE, Berkshire, UK Registered Office as above.

1 0

(no subject)
by Praveen Madala 15 Nov '04

15 Nov '04

Hi, I have been trying to MODEL a protein with distance restraints using MODELER version-12 with InsightII . Here I am attaching my .rsr file, and the below are the problems i got in my trials.... MODELLER12 VERSION: USER FORMAT R 3 1 1 21 2 2 0 7.1000 0.3387 SG:140 SG:222 R 3 1 1 21 2 2 0 16.0000 0.3387 CA:139 CA:248 R 3 1 1 22 2 2 0 17.0000 0.3387 CA:139 CA:248 R 3 1 1 21 2 2 0 17.0000 0.3387 CB:139 CB:248 R 3 1 1 22 2 2 0 18.0000 0.3387 CB:139 CB:248 R 3 1 1 21 2 2 0 16.0000 0.3387 CA:139 CA:251 R 3 1 1 22 2 2 0 17.0000 0.3387 CA:139 CA:251 R 3 1 1 21 2 2 0 16.5000 0.3387 CB:139 CB:251 R 3 1 1 22 2 2 0 17.5000 0.3387 CB:139 CB:251 R 3 1 1 21 2 2 0 19.6000 0.3387 CA:139 CA:252 R 3 1 1 22 2 2 0 20.6000 0.3387 CA:139 CA:252 R 3 1 1 21 2 2 0 21.5000 0.3387 CB:139 CB:252 R 3 1 1 22 2 2 0 22.5000 0.3387 CB:139 CB:252 R 3 1 1 21 2 2 0 13.3000 0.3387 CA:139 CA:249 R 3 1 1 22 2 2 0 15.3000 0.3387 CA:139 CA:249 R 3 1 1 21 2 2 0 11.4000 0.3387 CB:139 CB:249 R 3 1 1 22 2 2 0 13.4000 0.3387 CB:139 CB:249 R 3 1 1 21 2 2 0 15.0000 0.3387 CA:139 CA:250 R 3 1 1 22 2 2 0 17.0000 0.3387 CA:139 CA:250 R 3 1 1 21 2 2 0 15.3000 0.3387 CB:139 CB:250 R 3 1 1 22 2 2 0 17.3000 0.3387 CB:139 CB:250 R 3 1 1 21 2 2 0 10.0000 0.3387 CA:117 CA:292 R 3 1 1 21 2 2 0 6.00000 0.3387 N:117 N:292 R 3 1 1 27 2 2 0 8.00000 0.3387 CB:117 CG:208 R 3 1 1 27 2 2 0 10.0000 0.3387 CB:117 CG:208 R 3 1 1 27 2 2 0 8.00000 0.3387 CB:117 CG:211 R 3 1 1 27 2 2 0 10.0000 0.3387 CB:117 CG:211 1) Modeler is not recognising the "SG" (first line) or "CG" (last four lines) atoms, when i am trying to run the MODELER program. 2) If I remove these lines the program is getting CORE dumped. and also I didnt under stand when to use 21, 22, 27 and 30 in distance restraints. can you please advice me what might be the reason, and what should i do.... regards, Praveen

2 1

model a peptide
by aseem mishra 15 Nov '04

15 Nov '04

dear all, i wish to model a peptide with few unnatural amino acids in it. can some one please suggest a way. thank you. Yahoo! India Matrimony: Find your life partneronline.

2 1

Modeller vs Prime or MOE
by Stradella, Omar 13 Nov '04

13 Nov '04

Are there any experienced users of Modeller out there that have tried other homology modeling software like Prime from Schrodinger or MOE from the Chemical Computing Group ?. Experiences, pros and cons ? Thanks Omar Stradella This e-mail, including any attachments, is a confidential business communication, and may contain information that is confidential, proprietary and/or privileged. This e-mail is intended only for the individual(s) to whom it is addressed, and may not be saved, copied, printed, disclosed or used by anyone else. If you are not the(an) intended recipient, please immediately delete this e-mail from your computer system and notify the sender. Thank you.

2 1

(senza oggetto)
by Matthias Negri 10 Nov '04

10 Nov '04

I wondered whether it's possible to build a DNA complex model? (My model is a heterodimer, bounded as a dimer on the DNA!) I'm sure it is, but i can't figure out which protocol i have to define.. Any tip or hint would be very appreciated.. Thanks Matthias Negri Univesity of Bologna

1 0

basic example erros
by butheryin 05 Nov '04

05 Nov '04

Hi, Modeller users I had tried to run the example 1 in workshop of the web site.but it reported lots of erros as follows on p4 win2k sp2 patch : mod7v7 seq_search.top Date and time of compilation : 09/13/2004 17:17:22 Job starting time (YY/MM/DD HH:MM:SS): 2004/11/05 15:41:50.484 rdpir___223E> Array too small. Increase MAXRES current maximum, current need: 0 1 recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 4784113 4671.985 4.562 Starting time : 2004/11/05 15:41:50.484 Closing time : 2004/11/05 15:41:51.437 Total CPU time [seconds] : 0.95 mod7v7 align.top Date and time of compilation : 09/13/2004 17:17:22 Job starting time (YY/MM/DD HH:MM:SS): 2004/11/05 15:38:00.015 fullfn__230E> File not found: align.top Directories: :${MODINSTALL7v7}/bin\ Extensions : .top: Prefixes : nothing, pdb recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 56755 55.425 0.054 Starting time : 2004/11/05 15:38:00.015 Closing time : 2004/11/05 15:38:00.937 Total CPU time [seconds] : 0.92 mod7v7 TvLDH-4mdh.top Date and time of compilation : 09/13/2004 17:17:22 Job starting time (YY/MM/DD HH:MM:SS): 2004/11/05 15:40:15.625 fullfn__230E> File not found: TvLDH-4mdh.top Directories: :${MODINSTALL7v7}/bin\ Extensions : .top: Prefixes : nothing, pdb recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 56755 55.425 0.054 Starting time : 2004/11/05 15:40:15.625 Closing time : 2004/11/05 15:40:16.562 Total CPU time [seconds] : 0.94 mod7v7 model-single.top Date and time of compilation : 09/13/2004 17:17:22 Job starting time (YY/MM/DD HH:MM:SS): 2004/11/05 15:40:44.375 check_a_337E> Structure not read in: 1 recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 5436769 5309.345 5.185 Starting time : 2004/11/05 15:40:44.375 Closing time : 2004/11/05 15:40:45.625 Total CPU time [seconds] : 1.25

2 1