Figure 1.
Clinically used antimalarial chemotypes and their corresponding MoA.
Figure 2.
Contributions of the individual data sets to the predicted target space.
(A) Target space of active antimalarials only; (B) target space including both active and inactive compounds.
Figure 3.
Protein classes of predicted targets.
White bars represent the predicted relevant target space, whereas the different prioritized subsets are depicted by shaded bars.
Table 1.
List of predicted 39 high priority malaria targets and their TDR ranking.
Figure 4.
Drug-target network of 1,908 active compounds predicted for 147 P. falciparum proteins.
Node colors encode target families (red: kinase; orange: protease; yellow: other enzyme; green: ribosomal protein; magenta: transporter/channel; blue: other protein; grey: unknown function), node shapes encode prioritization (hexagon: high priority; triangle: high affinity target; diamond: enriched target; square: other predictions). Targets from the same orthologous group are merged into one common node. Compounds are shown as white circles and grouped according to their target profiles. Edge width corresponds to the number of compounds in the respective group. Clinically used drugs are highlighted in light blue. The numbering corresponds to chloroquine, mefloquine, and artemether (1), quinine (2), pyrimethamine (3), proguanil (4), sulfadoxin (5), and atovaquone (6). Capital letters are used to identify the target hubs of Hsp90 (A), plasmepsin I, II, IV, and VI (B), bifunctional dihydrofolate reductase-thymidylate synthase (C), acyl-CoA synthetase (D), serine/threonine protein kinase ARK2 (E), and falcipain 2a, 2b, and 3 (F).
Figure 5.
High priority targets versus total number of targets within a predicted profile.
The size of a circle relates to the number of compounds that address a profile of the given characteristics.
Figure 6.
Selection of diverse multi-target profiles.
The structures shown correspond to (A) the core scaffold of a set of 19 2,4-diaminoquinazolines, (B) TCMDC-132054, and (C) GNF-Pf-2272. The corresponding distributions of predicted affinities over multiple targets are provided on the right-hand side. For the set of 2,4-diaminoquinazolines (A), average predicted affinities are shown, with error bars giving the standard deviation over all compounds. Asterisks within the bars indicate the priority class of the respective target (‘****’ = high priority, ‘***’ = high affinity, ‘**’ = enriched, ‘*’ predicted target). Targets of the same orthologous group are joined to a single bar.
Figure 7.
Data work-flow and library sizes during pre-processing of virtual target profiling.