An Invitation to Open Innovation in Malaria Drug Discovery: 47 Quality Starting Points from the TCAMS

Abstract

In 2010, GlaxoSmithKline published the structures of 13533 chemical starting points for antimalarial lead identification. By using an agglomerative structural clustering technique followed by computational filters such as antimalarial activity, physicochemical properties, and dissimilarity to known antimalarial structures, we have identified 47 starting points for lead optimization. Their structures are provided. We invite potential collaborators to work with us to discover new clinical candidates.

Keywords: Malaria; TCAMS; lead optimization; open innovation.

Figure 1

Top panel shows P. falciparum 3D7 XC₅₀ vs total score (Total Score Fast-track) (see the text). Each spot represents a compound with the color of each compound/spot reflecting its potency with a gradient from most potent in red (≤0.5 μM) to least potent in blue (0.51–1 μM). Note that many compounds overlap resulting in black coloring. As might be expected, the compounds with the highest scores tend to be the most potent. The bottom panel shows those compounds scoring more than 15 with cluster number (Complete_Link_Clus) plotted vs score (Total Score Fast-track). The colors of the compounds reflect their potency with XC₅₀ ≤0.5 (red), 0.51–1 (blue), 1–1.5 (yellow), and 1.51–5 μM (black). Compounds appearing in the same column are members of the same cluster. Note that some clusters only contain one compound (singletons).

Figure 2

Most druggable/interesting structures from the 47 clusters that remain after the filtering process. Note that some of the scaffolds in these different series (clusters) are structurally very similar (e.g., series 16 and 24)—here, the algorithm has focused on the dissimilarity of the peripheral functionality.

Figure 2

Most druggable/interesting structures from the 47 clusters that remain after the filtering process. Note that some of the scaffolds in these different series (clusters) are structurally very similar (e.g., series 16 and 24)—here, the algorithm has focused on the dissimilarity of the peripheral functionality.

Figure 3

Plot showing the representatives of the 47 selected series comparing potency/clogP (red, cLogP ≤ 3; blue, clogP 3.1–5; yellow, clogP 5.1–6; and black, clogP > 6).

Figure 4

Exemplar compounds and their profiles for the top five selected series. For comparison, chloroquine has Pf IC₅₀ = 0.024 μM.

Figure 5

Analogues of TCMDC-125454 in the aryl carboxamide series found in TCAMS.