Rosetta Antibody cuts off light chain?

Member Site › Forums › Rosetta 3 › Rosetta 3 – General › Rosetta Antibody cuts off light chain?

Viewing 1 reply thread
  • Author
    Posts
    • May 2, 2019 at 10:17 pm #3132
      Anonymous

        Hi all,

        I am trying to use Rosetta antibody to predict the structure of the protein 5i19. Here is the input fasta:

        >heavy

        EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTYYADSVKGRFTISRDNSKNTLYL QMNSLRAEDTAVYYCAKYDGIYGELDFWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGAL TSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKS

        >light

        DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQ
        PEDFATYYCQQSYSTPLTFGQGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVT
        EQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

         

        For some reason, the template modeling process cuts off the light chain at position 109. Any thoughts?

        Thank you!

        Ziheng Wang

         

      • May 3, 2019 at 2:00 pm #14697
        Anonymous

          Dear Ziheng,

          RosettaAntibody only models the variable region of the antibody, for the light chain in your case this would be residues 1-109 (under the Chothia numbering convention). We do not consider the sebsequent residues as they comprise the constant domain, which tends to be well-conserved and not involved in antigen binding. However, if the constant region of this antibody is important for your modeling task, I’d be happy to help you add it to your model (it would involve something like modeling the variable domain alone, then finding a structural template for the constant domain, aligning the variable domain to the template, altering the template sequence in the constant domain to match the target, and remodeling the linker regions to connect the modeled variable and constant domains).

          Best,

          Jeliazko 

      • Author
        Posts
      Viewing 1 reply thread
      • You must be logged in to reply to this topic.