Hello, everyone I prepared paper_interface_design_pilot_commands. List when I did RIFdocking step 12 Running a pilot job. When I ran the commands, the error occurred:
basic.io.database: Database file opened: scoring/score_functions/sap_sasa_calib.dat
WARNING!! DALPHABALL output nan at index 140724834832176
DALPHABALL output indicies not matching! 1!=140724834832176
ERROR:: Exit from: src/core/scoring/packing/surf_vol.cc line: 170
protocols.rosetta_scripts.ParsedProtocol: [ ERROR ] Exception while processing procotol:
File: src/core/scoring/packing/surf_vol.cc:170
[ ERROR ] UtilityExitException
This is my design flag:
-script_vars patchdock_res=51,53,55,76,79,80,82,85,87
-script_vars runpsipred_single=/home/gaon/biosoftware/psipred/runpsipred_single
-dalphaball /home/gaon/biosoftware/rosetta39/main/source/external/DAlpahBall
-indexed_structure_store:fragment_store /home/gaon/cao_protocol/ss_grouped_vall_all.h5
#-dunbrack_prob_buried 0.8
#-dunbrack_prob_nonburied 0.8
#-dunbrack_prob_buried_semi 0.8
#-dunbrack_prob_nonburied_semi 0.8
I used Dalphaball rosetta3.9 version is because when I compiled Dalphball in rosetta3.13 version, it produced the multiple definition error. My gcc version is 11.4.
Has anyone encountered this error with Rosetta? How can I solve this problem? I'd be grateful if you could give me any help.
| Attachment | Size |
|---|---|
 ROSETTA_CRASH.log | 157.52 KB |
I'm not sure why the version with Rosetta 3.9 works for you but the one with Rosetta 3.13 doesn't. There shouldn't be all that many changes between the two. The only difference I can see when double checking is better support for larger proteins. (The one for Rosetta 3.9 is limited to 25000 atoms, or ~1500 residues.) -- I can certainly see the compiler version making a difference, though, as newer compilers may be pickier about things than older ones are.
One issue may be that you're giving the path to the DAlphaBall directory, rather than the full executable name. Try `-dalphaball /home/gaon/biosoftware/rosetta39/main/source/external/DAlpahBall/DAlphaBall.gcc` instead.
Hi! Thank you for your answer! I recompiled the Dalphaball and rosetta3.13 with gcc version 6.3. They were both compiled successfully. And I changed the path in design.flags to make it have the full path to the full executable name.
But when I run the command:
It reports error:
/home/gaon/biosoftware/rosetta313/main/source/bin/rosetta_scripts.hdf5.linuxgccrelease: error while loading shared libraries: libzlib.so: cannot open shared object file: No such file or directory
However , when I try to look for this file, I found it is in the rosetta 313:
-rwxrwxr-x 1 gaon gaon 15K 1月 10 08:19 /home/gaon/biosoftware/rosetta313/main/source/build/external/release/linux/6.2/64/x86/gcc/6.3/hdf5/libzlib.so
Is there a solution to solve this problem? I'd be grateful if you could give me any help.
I found the error is because I changed the directory name after compilation. So, this time I changed the dircetory name before the compilation and recompile the rosetta 3.13 and Dalphaball using GCC 6.3 version again. The compilation was successful and I got the expected output!
Just in case someone has encountered the same problem:
1、for the multiple definiction error during the compilation of Dalphball, could change the compiler to a lower version, for gcc version 11.4, the error could occur. But when I use gcc 6.3, the Dalphaball could be compiled successfully.
2、could use ./DAlphaBall.gcc for checking the Dalphaball executable
3、the design.flags need to provide the direct path to the DAlphBall.gcc and rosetta 3.13 is better for bigger protein
Thank you for your help, rmoretti! I could continue my protocol now！