dear RDKit experts,

i have a question about conformer generation, maybe best posed on an

i'm working with macrocycles which have few amide groups. As in each amide
group H(-N) and O(=C) atoms can have two positions with respect to a rigid
part of a molecule (amide groups are direct neighbours of aromatic rings) -
"syn" and "anti" - one can think of few families of conformers: "syn syn
syn...", "syn syn anti...", "syn anti anti...", etc.

i would like to generate a set which would have representatives of all/most
of these families;
for that:
1. i started with constructing 3D structures of all representatives
manually (i.e. starting from the crystal structure, manipulating atoms in
all amide groups and pre-optimizing with MM in Avogadro just to have a set
of candidates representing all orientation families mentioned above)
2. then used AllChem.EmbedMultipleConfs with all the 3D structures
generated in step 1 as initial geometries, followed by MM optimization with

But as explained in
the program starts with creating the "distance bound matrix" based on
empirical information including "ideal bond lengths, ideal bond angles, and
a few ideal torsion angles", therefore i'm wondering whether what i do
makes sense: does it matter with which conformer ("syn syn...", "syn
anti..." etc.) i start if anyway for all the conformers of the same
molecule the distance bound matrix would be the same? is that correct?

If that is indeed the case (= it does not matter with which 3D structure i
start), what would be your advice to assure that all orientation types
("syn syn...", "syn anti..." etc.) are in a final set?
Since there is a random distance matrix generation step, would you simply
repeat AllChem.EmbedMultipleConfs many times until you see all structures
you want to have in a set?
Or is there a smarter way to do it?

Thank you and best regards,
Check out the vibrant tech community on one of the world's most
engaging tech sites,!
Rdkit-discuss mailing list

Reply via email to