Thank you very much for trying the Indigo nodes. Following your request, we have added the following properties:
Number of aliphatic atoms/bonds
Number of carbon atoms
Number of heteroatoms (non-C, non-H)
They will be available in next nightly build (i.e. tomorrow)
The number of components was already there (called "Number of connected components"). As for pKa, LogP, LogD, and Polar Surface Area, the present version of the Indigo library can not calculate these, sorry, but we can think about that properties in the future.
Thank you again for your useful suggestions. We can add the "Number of aromatic rings" into the next build, but I am wondering whether it should be all rings or just SSSR (smallest set of smallest rings)? In other words, take naphtalene: should Indigo report 2 or 3 aromatic rings for it?
As for the HBAs, HBDs, and rotatable bonds: Indigo can not calculate them directly, but you can get the result by using the Substructure Match Counter node with a SMARTS string as a query. You just connect your molecule(s) source to the first input of the Substructure Match Counter node, while to the second one you should pass a query molecule which you can obtain in the following way:
Table Creator (enter your SMARTS expression there) -->
Molecule Type Cast (cast it to the SMARTS type) -->
Query Molecule to Indigo (parse the SMARTS) -->
(to the second slot of the Substructure Match Counter)
Moreover, you can calculate the TPSA this way! (you asked about the Polar Surface Area before -- is that what you want?). However, this is a bit more complicated as it involves summing the substructure match counters (multiplied by coefficients) for different SMARTS. Good news: you can download an example of TPSA calculation from the KNIME Example Flow Server, look there into 099_Community/07_Indigo/09907006_indigo-tpsa-example. There are other examples there, too.
Thanks for the detailed information, I shall try this out for RBs, HBA, HBD, and PSA. Thanks for the instructions.
In terms of counting number of rings,I completely overlooked that you could call Napthelene 3 rings from the 2 small rings and the one larger ring!! As a chemist, I would just call Napthelene 2 rings.
I hope this doesnt make the implementation of "the number of aromatic rings" too complicated to do.
Is it possible to have a minor alteration to the Molecule Properties node. Now we are spoilt for choice with many options to calculate, is it possible to have them grouped a little more logically such as:
I agree that having them ordered that way is much more comfortable. They follow that order in the new version (1.0.0.0000948) that is available on the nightly updates site.
Thanks for ordering the properties more logically, its much easier to find the desired property. I'd really appreciate keeping the Molecular Volume and Polarisability properties on a wish list for the future.
Can you add number of "hetero atoms in aliphatic ring", "htero atoms in aromatic ring". I am trying to filter carbocylic, heteroaromatic and heterocylcic[non-aromatic rings] using row filter node. But cannot do so since it picks het-atom count can occur outside a ring too.
It would be great if we could go to the level of N-heterocycle!
InsilicoConsulting: Done. The new properties will be available tomorrow morning in the next nightly build. You can, though, count occurrences of SMARTS: "[!#6;A;R]" and "[!#6;a;R]" in the current build to get the filters you need.