#93 SMILES parser doesn't deduce radical centres

[Stephen Pascoe]

Radicals can be deduced in SMILES strings from
hydrogen-deficient atoms. E.g.

CC[CH2] == ethyl radical
CCO[O] == ethyl peroxy radical

I've tested these on OpenBabel and Marvin and both read
the SMILES correctly. CDK adds hydrogens instead.

[footnote]
Openbabel has a good Wiki page on radical SMILES
extensions at
http://openbabel.sourceforge.net/wiki/Radicals_and_SMILES_extensions.

[Egon Willighagen]

Logged In: YES
user_id=25678

Radicals are not supported in SMILES, an sich. Try it on
depict [1]: you get an error. Funnily enough I read that OB
wiki page some days earlier, and I read it in that it really
was an extension. CDK does not support this extension. Yet
at least. Default SMILES does not support it, so making this
a feature request.

BTW, thanx for the detailed report!

[Egon Willighagen]

Logged In: YES
user_id=25678

I think I forgot something :)

1.http://www.daylight.com/daycgi/depict

[Egon Willighagen]

Logged In: YES
user_id=25678

See also: http://www.lirico.co.uk/wp/?p=14

[Egon Willighagen]

No hydrogens are added anymore, but neither does it seem to visually show radical information. Try both SMILES here: https://cdkdepict-openchem.rhcloud.com/depict.html

[John May]

Tricky... these would be possible but CC[CH] for example, is that a singlet or triplet?

[John May]

Using CXSMILES:
CC[CH2] |^2:1|
CC[CH] |^2:2|
CC[CH] |^2:3|