Assessing the general applicability of chemical reactions

Interactive dashboard for running a standardized substrate selection workflow.

The procedure is straightforward and can be implemented as follows:

1. Gather a broad list of candidate substrates from commercial databases (e.g., Reaxys ©) or supplier catalogues.
   Alternatively, compile a list of your own substrates that you want to assess with respect to their chemical space coverage.
2. Filter the substrates for functional groups or substructures based on the reaction compatibility.
3. Go to "Generate / Assess Scope", upload your list and choose a UMAP model for dimensionality reduction.
   There is a pretrained, optimized model available.
4. Select the number of clusters and the number suggestions.
   Additional filters on molecular weight range, removal of inorganic compounds and desired substructures (SMARTS pattern) can also be specified.
   Finally, click on submit to generate the representative set of standardized substrates.
5. The selected substrates can be downloaded as an Excel or CSV file and visualized in the UMAP projected drug-substrate map to assess chemical space coverage.

For a more detailed guide on the web application or instructions how to implement the workflow using Python, refer to our supporting information