Re: [Rdkit-discuss] Morgan atom invariants in the atom-pairs fingerprint?

2016-06-24 Thread Greg Landrum
Hi Sereina, At the moment, for historical reasons, the atom pair and topological torsion fingerprint is limited to invariants that are at most 9 bits long (i.e. <512). I think it should be possible to remove the restriction, but it requires a bit of rethinking of the way the code works. What you

Re: [Rdkit-discuss] Morgan atom invariants in the atom-pairs fingerprint?

2016-06-24 Thread Nadine Schneider
Hi Sereina I had the same problem a while ago :). I think the reason is that the AP fingerprint invariant only allows hashes for atoms up to an integer value of 511 (9 bits) and the Morgan connectivity invariant returns 32 bit codes, therefore you get this error "larger than allowed value". What

[Rdkit-discuss] Morgan atom invariants in the atom-pairs fingerprint?

2016-06-24 Thread Sereina
Hi RDKitters, Is there a way to use the Morgan (connectivity) atom invariants in the atom pairs fingerprint? I tried naively the following: inv = AllChem.GetConnectivityInvariants(m) AllChem.GetHashedAtomPairFingerprintAsBitVect(m, atomInvariants=inv) but it gives me the error “ValueError: