Binding site orientation in Composite

[ab9010]

Is there any straightforward way to invert protein binding site on DNA? Since many DNA-binding protein complexes are not symmetrical, nor their recognition sequence on DNA is symmetrical, I noticed that the tool automatically interprets the 'binding site' from the reference PDB in just one of the two possible orientations. Any clue on how to select the second orientation of the binding site to make the model through Composite?

e.g.  my seq:  aaaaaaATTGGaaaaaaaaaaaGAGATaaaaaaaa  (uperrcase are the two binding sites)

     ref pdbs:        aaaATTGGaaa           aaATCTCaaa
                   protein1 (correct orientation)    protein2 (the tool selects the complementary chain as reference, so the protein is also inverted in the final model)

Thank you in advance

[xiangjun]

Hi,

Thanks for your involvement on the 3DNA Forum.

I understand your question conceptually. In the current implementation of the "Composite" module of web 3DNA 2.0, however, there is no "straightforward way" to control the orientation of reference DNA-protein complexes. In principle, this feature should not be hard to implement, given the infrastructure already in place.

The "Composite" module was initially implemented by Guohui Zheng in "Web 3DNA—a web server for the analysis, reconstruction, and visualization of three-dimensional nucleic-acid structures" in 2009, and then polished by Shuxiang Li in "Web 3DNA 2.0 for the analysis, visualization, and modeling of 3D nucleic acid structures" in 2019. The initial w3DNA website hosted at Rutgers (http://w3dna.rutgers.edu) is no longer functioning. The w3DNA 2.0 website (http://web.x3dna.org) is hosted at Columbia to which I am direct access.

Shuxiang no longer works on the 3DNA project, even though he may still be available for quick answers. However, implementations of new features like this one are not expected from him. Within my capability, I will try to ensure that the web 3DNA 2.0 server works as described in the 2019 publication in Nucleic Acids Research.

Best regards,

Xiang-Jun

PS. Given the many recent and previous questions like yours, I am planning to distill and streamline related modeling utilities in 3DNA v2.4 into DSSR. The w3DNA 2.0 and wDSSR may not be updated in sync, but the DSSR command-line interface will be far more sophisticated. In due course, I may create a new simplified web interface (like http://skmatic.x3dna.org) specifically for the modeling capabilities in DSSR 2.0. 

[xiangjun]

Hi,

I have created a brand-new program for template-based modeling of nucleic acid structures, including DNA-protein complexes. Now users can easily specify their own templates, as well as their orientations, via the command-line. This novel modeling program is distributed as a module of DSSR Pro.

Best regards,

Xiang-Jun