Curated monomers for Virtual Forward Synthesis (VFS)

This data set (named "monomers.pkl") of molecules was used to generate the hypothetical polymers stored in https://github.com/Ramprasad-Group/polyverse-structures. This data set is derived from the union of the ZINC15 and ChEMBL databases. The data set was created using the following rules: (1) select the molecules that are readily purchasable (2) select the molecules that have standard reactivity (3) strip counterions and chirality (4) Use the SAscore to remove complex molecules.

The data set contains 7,754,881 entries. It contains two columns: "index" and "id-smiles". Each entry in "index" is an integer, starting from 0. Each entry in "id-smiles" is a tuple. The first element of the tuple is an ID (either the ZINC15 ID or the ChEMBL ID). The second element of the tuple is the processed SMILES string.