Hello Avogadro I had used the Avogadro (with gold lace) application to generate a 3D File from a PubChem 2D file, the ligand was large 89-heavy atoms/1000 Daltons, was Avogadro able to accurately 3D build this large of a ligand? Thanks if you know:)
There’s no “maximum size” and certainly I’ve converted molecules larger than 80-90 atoms. Do you have a particular file? Most are available already in 3D format.
Not sure what you mean by the “gold lace” though.
Hello ghutchis thank you for your kind update:) and yes my molecule I could not discover on PubChem nor PDB, regardlessly attached is the PubChem 2D Structural File that I had built into a 3D File via Avogadro (see screenshot for color as opposed to blue color). I am unsure how accurate the Avogadro final 3D conformation is, if you have any further feedback feel free to reply (I am to dock this ligand into a transmembrane protein, accordingly in search of a starting accurate conformation.). Thanks:)
10Panx1_PubChem_2D.sdf (20.3 KB)
You should use the latest version of Avogadro2: Install — Avogadro 1.102.1 documentation - yes it’s blue with a little number 2 in the icon. I can see that the current builds of Avo2 don’t recognize that the file is 2D only and needs 3D coordinate generation.
The easy workaround is to Edit ⇒ Copy As ⇒ SMILES and then in a new file, Build ⇒ Insert ⇒SMILES
I’m not sure I understand what you mean by “how accurate” a 3D conformation is.
A molecule like this is large and flexible. There are likely to be multiple thermally accessible conformers, so you should generate several. There is no one measure of “accurate” for a case like this.
For example, we discussed this with my colleague David Koes in a recent paper:
“Conformer Generation for Structure-Based Drug Design: How Many and How Good?”
DOI: 10.1021/acs.jcim.3c01245
You can use Avogadro2 to generate starting conformers. For example - you can use the Genetic Algorithm / RSMD scoring method:
Or you can use RDKit or other tools as discussed in the paper.
And here’s your file, by the way:
example.sdf (15.2 KB)
Hello ghutchis thank you for your kind update:), understood, agreed and this has been my methodology i.e. to take more than one conformation towards docking. Also thank you for the paper (I will read it) and the file!
1 Like
For what it’s worth, Avogadro2 now has this patch building, which will use the 3D coordinate generation from Open Babel when reading 2D SDF files:
Hi ghutchis thank you for your kind update:) and I had Avogadro (Gold) 1.2.0 Generated a nice 3D Coordinate File for the ligand indicated in this thread via a PubChem 2D.sdf File (nice meaning regardless of the multiple conformations, this conformation matched similar ligand experimental visual bindings) that I selected towards docking, accordingly would the Avogadro2 generate something substantially more accurate? (I inquire since I had already docked this and selected the best model.) Thanks if you know:)
No, both versions of Avogadro use Open Babel for coordinate generation and conformer searching.
Hi ok thanks a bunch for the affirmation:)