Hi #CompChem community, have a question on datasets. I've observed that many open-source #ADMET_datasets contain a decent % of peptide-like and macrocycle-like cpds. Would you consider filtering out these cpds to develop models specifically tailored for small molecules?.