1440 suzuki rxns

[bdeadman]

1440 suzuki rxns from https://doi.org/10.1038/s44160-023-00351-1.

[bdeadman]

@cernak-lab This is the 1440 rxn Suzuki dataset from your Miniaturization of Reactions paper (https://doi.org/10.1038/s44160-023-00351-1.

suzuki for enumeration.zip contains the .pbtxt template and the suzuki sheet of rxns in csv format.

I'm working on the reductive amination and alkylation datasets in the same paper.

[bdeadman]

This PR now has all 3 datasets [Suzuki, reductive amination, N-alkylation with BOC deprotection]. The data and template files are in submission.zip.

Notes:

• The alkylation data has 2 templates and data spreadsheets which were enumerated separately (a subset had a different template pattern), and then the datasets merged.
• I have queried some of the product labels/assignments in the alkylation data with Tim. If we get more information I will update this part of the alkylation dataset.

[bdeadman]

ValueError: error parsing data/alkylation_merge.pbtxt: 2:331 : Message type "ord.Dataset" has no field named "BOC".

Looks like I've accidentally broken them template file at a "BOC" string in the alkylation dataset. I'll find the error and update the file(s).

[bdeadman]

Fixed the "BOC" string problem (it was a rogue " in the dataset description). Now having issues with rdkit validation of a product SMILES (Core_SMILES). It is odd since it was validating when I enumerated the datasets.

[bdeadman]

The Core_SMILES not validating was due to the tag on the SMILES identifier not having the $$ labels. This is corrected in the attached template file, and in the GitHub PR.

alkylation_template.zip