An Open Drug Discovery Competition: Experimental Validation of Predictive Models in a Series of Novel Antimalarials

Abstract

The Open Source Malaria (OSM) consortium is developing compounds that kill the human malaria parasite, Plasmodium falciparum, by targeting PfATP4, an essential ion pump on the parasite surface. The structure of PfATP4 has not been determined. Here, we describe a public competition created to develop a predictive model for the identification of PfATP4 inhibitors, thereby reducing project costs associated with the synthesis of inactive compounds. Competition participants could see all entries as they were submitted. In the final round, featuring private sector entrants specializing in machine learning methods, the best-performing models were used to predict novel inhibitors, of which several were synthesized and evaluated against the parasite. Half possessed biological activity, with one featuring a motif that the human chemists familiar with this series would have dismissed as “ill-advised”. Since all data and participant interactions remain in the public domain, this research project “lives” and may be improved by others.

Optimisation-based modelling for drug discovery in malaria

“…The piecewise equations in Table 1 suggest that the Series 3 compounds are not likely to be active, as the maximum activity values calculated from the two equations are either pIC 50 = 5.47 (when Bit 0484 = 1) or pIC 50 = 4.60 (when Bit 0484 = 0 and Bit 0179 = 1), which is lower than our defined threshold of pIC 50 ≥ 5.80. 19 This is also supported by the distribution of the true pIC 50 values of Series 3 compounds. As shown in Figure S3c, the median and mode value of Series 3 are both less than 5, and only 2 out of 49 compounds are considered to be active, namely OSM-S-106 and OSM-S-590.…”

“…On the other hand, if a compound includes this fragment (Bit 0350 = 1), the model predicts that its bioactivity will be pIC 50 = 4.30, thus inactive against Plasmodium Falciparum , if an activity threshold pIC 50 ≥ 5.80 is assumed. 19 By exploring m01 (Region 01) equation, we can also see which of the remaining fragments selected by the algorithm make positive or negative contributions to the bioactivity of these compounds.…”

“…This study focused first on building a predictive model for Series 4 analogues, 19 but as our method can inherently distinguish structurally heterogeneous chemical sets, all series and assays were then considered for a more comprehensive analysis. A raw dataset containing all OSM compounds from Series 1 to 4 and their respective assay data was downloaded from the Master List of chemicals provided by OSM.…”

“…Recently, a competition was launched to develop and evaluate strategies for accurate prediction of anti-PfATP4 activity among Series 4 compounds from the OSM master chemical list database, thereby reducing project costs associated with the unnecessary synthesis of inactive compounds. 19…”

“…In this study, we build upon previous work on development of an interpretable model for QSAR modelling based on mathematical optimisation, modSAR, 20 and model the inhibition activity ( pIC 50 ) of compounds from the OSM dataset against Plasmodium falciparum . We have previously applied modSAR to an earlier version of the OSM data using pre-defined molecular descriptors, 19 and here we develop this work further in order to offer a better understanding of these antimalarial candidates as well as paving the way for future SAR explorations, lead optimisation and new de novo drug design efforts for malaria.…”