How to perform a DNA structural alignment in pymol

Question

How can I "fit" two DNA structures having different nucleotide sequences in PyMol?

I would like to use the structure of a DNA-binding protein in PDB (1h9t), which is bound to DNA in the PDB file, and fit a longer DNA built in silico with a 3D dart (haddock.science.uu.nl/dna/dna.php).

Answer 1

Aligning in PyMol

PDB ID 1h9t contains a structure of both a DNA and a protein. It starts as one object.

Before we perform an alignment, we need to separate your DNA from the protein.

1. Show sequence (click the S at the bottom right).

2. Highlight all the "residues" from the chains X and Y (these chains contain each strand of your DNA) in the sequence bar at the top.
3. Use the GUI to extract selection to new object (it's a faff doing this by terminal I find).
4. Rename new object (automatically named obj01 probably) to DNA1, or whatever you prefer.
5. Load your new, unaligned piece of DNA (I used fetch 1BNA).

6. Pymol uses the CEalign algorithm to align structures in 3D space. You can call it using the following command:

   cealign DNA1, 1BNA, object=aln

Output:

Another 3D alignment software I really like is the Dali server.

Haddock

Haddock doesn't care if two molecules are aligned pre-analysis assuming you are submitting each DNA separately alongside a protein for binding analysis. This is because it tries a huge variety of orientations. Here is a list of things that you do need to do to your pdb before submitting to HADDOCK