Standardization, SMILES format and CoA issues in Reactor

Some questions received in email:





1. In our evaluation copy of Reactor, it looks like you can apply the Standardizer procedures to a reaction scheme using "Advanced Options -> Product Standardization". Is this a standard feature of Reactor, or is this only available with Reactor Pro? To use these options is it necessary to have a license for Standardizer?





2. If we have a list of compounds in a SMILES format that was calculated by another program, how would you recommend converting the compounds to the ChemAxon SMILES format? When we store the compounds in our database, we store the SMILES string and it needs to be calculated in a consistent way (i.e. from the same algorithm). I know Standardizer can do this, but is there another way? Again, we would like to minimize the number of tools we have to buy.





3. I asked this in my last email but didn't clarify: how difficult would it be to implement the !match(CoA) function *without* an atom map? Is it possible for me to make a function like this myself, or is a developer's help needed?

1. Yes, Standardizer license is needed. However that will be included in the Reactor Pro version since the upcoming JChem 5 release (limited to use within Reactor).





Let me briefly explain some technical issues: there are two points in Reactor to include standardization.


- On the one hand, each reaction can convert the input reactants to the expected format. There is a Standardization tab in the Reacton Editor for this purpose. Imagine a reaction scheme, that reduces nitro groups. A simple nitro transform can be added to the reaction to convert the ionic nitro mesomers to the neutral form expected by the reaction scheme. The same method can be used for example to convert oxo compounds to enols if the scheme expects enols as starting compounds.





- On the other hand, the resulting products of a reaction can be beautified (ie. dearomatize, clean) or simplified (remove stereo) with standardization. That is available in the "Advanced Options" of the Reactor wizard.





In short, reactant standardization is included in the generic reaction, while product standardization is part of the enumeration process.








2. File conversion is really easy with Standardizer, select input files (even is different formats), do not specify any actions, but specify output in the expected format.





Standardizer, however, is not the only tool available for this purpose, no need for Standardizer license for SMILES conversions.Marvin View for example, can save molecules in SMILES. If you are familiar with command line, the most effective tool is MolConverter.








3. You would like to prohibit reaction in a CoA linked molecule. It is possible if you add a substructure matching expression to the reactivity rule of that reaction:

Code:

!match(ratom(2), "CC(C)(COP(O)(=O)OP(O)(=O)OCC1OC(C([OH1X2:1])C1OP(O)(O)=O)N1C=NC2=C1N=CN=C2N)C([OH1X2:2])C(=O)NCCC(=O)NCCS", 1, 2)

It works, but very difficult to undestand without a description. So I added CoA to the list of ligands known by Chemical Terms, it will appear in JChem 5, so you do not need to type the SMARTS of CoA in the Chemical Terms expressions, its name is enough. From JChem 5, this will be a valid expression

Code:

!match(ratom(2), CoA, 1)

The CoA in the dictionary has atom map 1 on the sulfur.








We plan to simplify the syntax further, similar to the format you propose (no atom maps):

Code:

!match(ratom(), CoA)

Unfortunately, this modification needs some development in the Chemical Terms parser and evaluator code. I think, that our developers have to work on it.

Are the standardization features you described also available in Reactor Standard?

The features are available, but require a separate Standardizer license. In case of Reactor Pro, no separate Standardizer license is required. Having the Pro version you are licensed to use Standardizer and all calculation plugins within Reactor with no extra fee.