Electrophilic motifs and thiol reactivity in natural-product chemical space: data, analysis code and quantum-chemical records

This record holds the data and the analysis code behind a cheminformatics study of wherecovalent, thiol-reactive chemistry sits in natural-product structural space, and how itsdistribution differs from that of synthetic compounds with recorded biological activity. A library of 22 SMARTS motif definitions and a single site-enumeration procedure were applied,unchanged, to three collections: 11,424 standardised records of an African natural-productdatabase, 695,129 records of a global natural-product collection, and a size-matched sample of695,130 bioactive synthetic structures. Every match is resolved to the individual reacting atom. Results carried in this record: - Motif prevalence in each collection, with absolute structure counts and 95% intervals on the ratio between the natural-product and synthetic sets.- Conjugated-carbonyl site density normalised by heavy-atom count, reported overall and within bands of molecular size, so that a size effect can be separated from a compositional one. Natural products carry 0.82 sites per hundred heavy atoms against 0.34 for the synthetic set.- Radius-2 reactive-environment vocabularies: 18,821 distinct environments over 203,833 natural-product sites and 11,607 over 69,313 synthetic sites, of which 3,129 are common to both, together with the share of enumerated sites those common environments carry.- A thiolate-addition energy computed with GFN2-xTB under an implicit water model on truncated local fragments at 2,993 mapped sites, calibrated per reaction family against 36 glutathione rate constants taken from the literature, with 598 site predictions released under a stated applicability domain.- Cross-validation, response-permutation and external-validation statistics, and three graph-derived fragment-fidelity diagnostics tested by rebuilding sites at a wider truncation radius.