You are here

Rosetta_CM error unrecognized atom 'CS' this atom is not in PDB

4 posts / 0 new
Last post
Rosetta_CM error unrecognized atom 'CS' this atom is not in PDB

Hi all,


       I'm sorry if this is an old thread, I did find other forums dealing with unrecognized atoms but in those threads my problem was not resolved.  I am trying to follow the rosetta_CM tutorial and I got through the script and it generated the .xml file which I checked and it looked good. However, when I run the hybridized steps using rosetta scripts I get this error:


ERROR: unrecognized atom_type_name 'CS'

The unfortunate thing is I can't find any CS atoms or CS (using the find function in vim) in my PDB

The second line says this:

File: src/core/chemical/

I'm pretty novice at coding but I tried to follow it and there is a line that says 'ignore white space' with a code line that looks like it removes spaces. I had an idea that maybe it trimmed wo much and combined a C with a S from a SER which I do have in my PDB but this is very unlikely as I've done quite a bit of modeling without this issue. Although this is my first time using 3.10, the last time I used rosetta was with 3.6.


I've attached the PDB that the setup python script created and the spacing looks normal and in line. Please let me know if you need more information to try to troubleshoot this problem.



sk4T2.pdb105.9 KB
Post Situation: 
Thu, 2019-01-24 19:43

I think this is due to a binary/database version mismatch.

There's a rather tight coupling of the version of the Rosetta executables with the version of the database. You should only every use the version of the database which comes with the version of the compiled code. (For example, if you're using the executables compiled from Rosetta 3.9, you should only ever use the database which came with Rosetta3.9 -- The database from Rosetta3.8 or the one from Rosetta3.10 is not necessarily going to work.)

My guess is that is what's happening here. You're using one version of the Rosetta executable, but the -database option in your command line or in your options file is specifying the database from a different version. Alternatively, it might be that you have the ROSETTA3_DB environment variable set, and it's set to a database version which is not the same as the executable you're using.  (Note that an explicitly provided -database option will override any ROSETTA3_DB setting.)

Fri, 2019-03-29 12:50


I think this has to do with the input required for the score function. If you are using a score function that needs centroid representation (cen), you have to add either in:file:centroid_input in the command line,  or the mover "<SuitchResidueTypeSetMover name="tocen" set="centroid"/> in your script.

I hope this helps.

Best regards,

Karen Gonzalez


Fri, 2019-05-31 07:18

For users implementing the beta score function, another possibility is a score function mismatch between input file and RosettaScripts XML. For example, a RosettaScripts XML specifiying the beta score function would also need the -beta flag specified as command-line input. 

Sun, 2020-08-23 14:03