I am using Avogadro 1.2 and am interested in being able to generate ssDNA and RNA molecules however when I try to build the molecules the structure is not what is predict from other platforms. For example, when I use the sequence 5’- CGGATGGAATTTTGTTCGTTGATTCCATCCG -'3 in the Build DNA/RNA application under the Build, I get a straight sequence. In reality, I should get a loop structure that has folded back on itself. I have not been able to generate these correct structures in Avogadro. Does anyone have any experience with this? Or tips?
The following is based on the observation that Avogadro’s file import/export so often interacts with OpenBabel. Perhaps a bit naïve.
except the leading 5'- and the trailing -'3 you mention, I pasted string CGGATGGAATTTTGTTCGTTGATTCCATCCG into a file test_fasta.txt, because I assume it is Fasta notation.
from the CLI, the content was converted into the .xyz format – which only reports the atoms, and their coordinates – by the command
obabel -ifasta test_fasta.txt -O test.xyz
Avogadro reads the content of then created file test.xyz as a double helix. The connections drawn look like single bonds.
With Jmol, the test.xyz equally is displayed as a double helix. Should the automatic computation of bonds be switched off (edit → preferences → tab bonds), you may issue the command connect; on Jmol’s console (File → Console), too.
At the moment, there are no tools in Avogadro to build loops. As @Thomas mentioned, you can construct a variety of helix motifs SS, DS, etc. but not in the context of a loop or origami.
That said, in Avogadro v2, there’s a robust system for plugins to interact with other programs or websites e.g. RNAfold, etc.
You can certainly read PDB and other files into Avogadro after generation.
Thank you. I am not an expert in OpenBabel but in the input FASTA file can I add the secondary structure and then get the coordinates from that structure? If so, this might work for my purposes.
Is there documentation for Avogadro v2 available? I went to the website but it said the documentation was not available. I currently use RNAfold for the secondary structure and if it I can link that to generate the 3D coordinates that would be ideal.
I’m not aware this were possible. Reading OpenBabel’s documentation about FASTA (the first of the data formats relevant to biology, section 18.16.1 in the pdf, or on-line (here)), there is no documented approach for this.
On the other hand, I infer similarities in what you aim for with in silico protein folding when results from X-ray crystallography and 3D-NMR spectroscopy complement knowledge about the sequence of the amino acids in chains, sheets, loops, barrels, etc. Thus, I suggest to address your question on mattermodelling.se. (For future users of Avogadro reading this thread, it would be helpful if you would add a link here to your thread in the other forum.)