You are here

Modeling the structure of camel single domain antibody

2 posts / 0 new
Last post
Modeling the structure of camel single domain antibody

I am now interested in modeling the structure of camel single domain antibody and find that an algorithm was implemented a algorithm in Rosetta in 2011  ( The "Document for CAmelid application" ( mentioned some file for running the protocol, e.g.:








I have installed rosetta_src_2017.08.59291_bundle in my computer. However, the above files don't exit in this Rosetta package. Could you tell me where I can find them?


Then I turned to the Rosetta text directory. There is a "antibody_H3_camelid directory" in the "rosetta/main/rosetta_tests/integration/tests" path. I ran the follow command in this "antibody_H3_camelid directory" directory:


      antibody_H3.linuxgccrelease @flags -database /app/rosetta_src_2017.08.59291_bundle/main/database/ -run:constant_seed -nodelay


and it generated a FR02_H_0001.pdb file. Is it the fine modeling result for the input sequence? There is a FR02_H.pdb file in the "input" subdirectory of the "antibody_H3_camelid directory" directory, so I guess the FR02_H_0001.pdb is more likely a relaxation or minimization result of the FR02_H.pdb .  Without a detailed explanation or tutorial, I can hardly understand how to run the test. Could you provide me with one such tutorial?


I would be very grateful if you could help.

Post Situation: 
Mon, 2018-01-08 00:34

The following is the reply from Jeliazko:

I'd be glad to help. I'm in the process of updating our camelid modeling protocol, so there is no concrete manual. So, to get started, I would recommend you download a version from Dec. 2017 -- it includes an update that permits camelid modeling and a crucial bug fix. Next, compile it (./scons -j<# of processors> mode=release bin), which should generate the antibody executable in $ROSETTA/main/source/bin/. Alternatively, you may be able to download a version with pre-compiled executables. Now, you can homology model camelid antibodies by supplying a fasta file with two lines: the first should read ">heavy" and the second should be the amino acid sequence. The command line to run modeling is: "/path/to/antibody -vhh_only -antibody:n_multi_templates 1". After about 20 minutes of runtime, this should result in a homology model where a homologous FR and homologous CDRs were found by BLAST search against a database and assembled. The final step is to run de novo modeling of the CDR H3 loop. This is done with the antibody_H3 executable and instructions on how to run this are analogous to the ones reported here:


Let me know if anything is unclear or you need more help. Camelid antibody modeling with the latest version of Rosetta is still under active development.



Mon, 2018-01-08 00:55