fix oxDNA export bug with incorrect helix coordinates in 3D

[zoombya]

Hi guys, great thanks for your work.
I have a great problem with exporting structures from scadnano to oxDNA.

1. any 3D structure designed in scadnano if I directly export to oxDNA get's flattened regardless of the lattice.
2. if I first export the structure to candnano2 and import using taco -> the structures look turned inside out.
   Attached is a 12hb exhibiting all these issues.
   And screenshots of the 2 mentioned export options.
   220921_0845_12hb_1kb.zip

[dave-doty]

Hmm, I can reproduce this behavior. Not sure what's going on.

I tried exporting with the Python package and it seems to work:

So I suggest that as a workaround for you for now. Install the scadnano Python package using the instructions here: https://github.com/UC-Davis-molecular-computing/scadnano-python-package#installation

Then put this code in a file named export.py:

import scadnano as sc


def main() -> None:
    design = sc.Design.from_scadnano_file('220921_0845_12hb_1kb.sc')
    design.write_oxdna_files()


if __name__ == '__main__':
    main()

FInally, run python export.py at the command line.

It will generate two files export.dat and export.top.

[zoombya]

great thanks, Dave.
It happens in both Hex and SQ lattices.

[dave-doty]

Here's a minimal example of the problem. It appears 3D coordinates are simply being calculated incorrectly, possibly by the function util.grid_position_to_position3d. This also fails to convert the square grid to the none grid:

In square grid:

After converting to none grid:

In the above image, helices 1 and 2 are on top of each other.