Fig. 1: Compound summary.

a Lipinski properties. Lipinski’s properties of compounds used in this study were calculated using StarDrop version 6.6.4. Left Y-axis: MW molecular weight: Right Y-axis: HBD hydrogen bond donor, HBA hydrogen bond acceptor, logD, logS. * indicates 80 compounds that yielded resistant clones. b Maximum Common Substructure (MCS). Structure similarity clustering analysis for 80 compounds yielding resistant clones and larger library of 1600, using Tanimoto as the similarity metric. The diagram shows 41 clusters sharing an MCS from which at least one compound was selected for drug response (indicated by diamonds). Circles represent compounds that were not selected, or inactive. The strength of cytotoxicity against the S. cerevisiae GM strain of tested compounds is indicated by the node’s color intensity from purple (higher potency) to yellow (lower potency). Probability values were calculated using the hypergeometric mean function showing that enrichment for clusters was greater than expected by chance. Compounds from clusters with a p-value of less than 0.05 and which had multiple members active against GM are shown. c Coding region mutations for selected compounds. Histogram showing the distribution of the number of coding mutations (e.g., missense, start-lost) per clone for the set of 80 compounds used in selections. d Gene enrichment for selected compounds. The p-value is the probability of repeatedly discovering the same gene for a given compound, calculated using Bonferroni-corrected hypergeometric mean function as described in Methods. Compound/gene pairs for n = 1, 2, and 3 can be obtained from Supplementary Data 4.