Table 1. Summary of E-Flux predicted mycolic acid inhibitors from Boshoff data set.

Predicted Inhibitors
Isoniazid	Specific	Strong	Known
Thiolactomycin	Specific	Strong	Known
Ethionamide	Specific	Strong	Known
ZnSO4	Non-specific	Strong	New
Ethambutol	Specific‡	Weak‡	Known
Cerulenin	Specific‡	Weak‡	Known
PA-21	Specific‡	Weak	New
Streptomycin	Non-specific	Weak	New
Valinomycin	Non-specific	Weak	New
Amikacin	Non-specific	Weak‡	New
Pyrazinamide * ‡	Non-specific	Weak	Known
Tetracycline	Non-specific	Weak	New
Dubos-NRP1; Dubos-NRP1+KNO3	Non-specific	Weak*	New
PA-824	Non-specific	Very weak	Known
Chlorpromazine‡	Non-specific	Very weak	New
Capreomycin	Non-specific	Very weak	New
Synthetic pyridoacridine analog (124196)‡	Specific	Very weak	New
Ascedidemin (111895)‡	Non-specific	Very weak	New
Rifapentine‡	Non-specific	Very weak	New
Procept 6776, 6778‡	Non-specific	Very weak	New
Succinate, palmitate in minimal medium*	Non-specific	Very weak	New
Starvation conditions	Non-specific	Very weak	Expected

Results are shown for all significant predictions from the set of 437 experiments in Boshoff et al [17]. Of the seven known inhibitors of fatty acid biosynthesis, E-Flux correctly predicts 6. Strong effects are defined as effects showing greater than +/−3 log change between control and drug; weak inhibitory effects have inhibition less than −1.5 log change and very weak effects less than −1. Weak enhancing effects have greater than +1 log change. Specific effects indicate effects on mycolic acid biosynthesis as contrasted to effects over a broad range of pathways – see text for more details.

^*Prediction made only for certain replicates.

^‡Prediction made only for certain doses.

Starvation conditions were phosphate- or Tris-buffered saline containing 0.05% Tween 80 (PBST or TBST) [17]. Conditions with both enhancing and inhibiting predictions among replicates were excluded.