A systematic and prospectively validated approach for identifying synergistic drug combinations against malaria

Abstract

BackgroundNearly half of the world’s population (3.2 billion people) were at risk of malaria in 2015, and resistance to current therapies is a major concern. While the standard of care includes drug combinations, there is a pressing need to identify new combinations that can bypass current resistance mechanisms. In the work presented here, a combined transcriptional drug repositioning/discovery and machine learning approach is proposed.MethodsThe integrated approach utilizes gene expression data from patient-derived samples, in combination with large-scale anti-malarial combination screening data, to predict synergistic compound combinations for three Plasmodium falciparum strains (3D7, DD2 and HB3). Both single compounds and combinations predicted to be active were prospectively tested in experiment.ResultsOne of the predicted single agents, apicidin, was active with the AC50 values of 74.9, 84.1 and 74.9 nM in 3D7, DD2 and HB3 P. falciparum strains while its maximal safe plasma concentration in human is 547.6 ± 136.6 nM. Apicidin at the safe dose of 500 nM kills on average 97% of the parasite. The synergy prediction algorithm exhibited overall precision and recall of 83.5 and 65.1% for mild-to-strong, 48.8 and 75.5% for moderate-to-strong and 12.0 and 62.7% for strong synergies. Some of the prospectively predicted combinations, such as tacrolimus-hydroxyzine and raloxifene-thioridazine, exhibited significant synergy across the three P. falciparum strains included in the study.ConclusionsSystematic approaches can play an important role in accelerating discovering novel combinational therapies for malaria as it enables selecting novel synergistic compound pairs in a more informed and cost-effective manner.Electronic supplementary materialThe online version of this article (10.1186/s12936-018-2294-5) contains supplementary material, which is available to authorized users.

Using Machine Learning to Predict Synergistic Antimalarial Compound Combinations With Novel Structures

“…The library of 23 compounds that were selected for prospective validation resulted from predictions generated by a developmental version of CoSynE that had previously virtually screened 21 million DrugBank combinations using the same training data, alongside a different approach that was developed in parallel to CoSynE (KalantarMotamedi et al, 2018 ; see Experimental section for details). From this library of 23 compounds (and a possible 253 combinations), a total of 20 combinations comprising 12 distinct individual compounds were predicted to be synergistic, and these were submitted for prospective experimental validation.…”

“…The approach by KalantarMotamedi et al ( 2018 ) differs from that described in this work primarily by the usage of gene expression data. Firstly, differential gene expression profiles of mild vs. severe malaria patient peripheral blood samples were used to predict potentially active single antimalarial agents by comparison of drug gene perturbations through a modified Gene Set Enrichment Analysis (GSEA) approach (Subramanian et al, 2005 ) applied to the Library of INtegrated Cellular Signatures (LINCS) Phase I database (Subramanian et al, 2017 ).…”

“…This selection was achieved by manually reviewing the top-ranked combinations (sorted by the probability of being synergistic that was assigned to each combination by CoSynE), and taking into consideration the prevalence of each compound throughout the list of combination predictions, followed by examining the literature co-occurrence of each predicted combination's compounds together with mention of P. falciparum in PubMed. These 10 chosen combinations comprised 18 compounds, and were submitted for testing together with an additional 10 selected from a different approach developed in parallel by KalantarMotamedi et al ( 2018 ).…”

“…Computational approaches have been investigated as a means to predict the synergistic interaction of compounds previously, with methods that utilize networks of pathways and simulation (Lehár et al, 2007 ; Nelander et al, 2008 ; Miller et al, 2013 ; Huang et al, 2014 ; Patel et al, 2014 ; Zhang et al, 2014 ), relationships between physicochemical properties (Yilancioglu et al, 2014 ), chemogenomics approaches (Bansal et al, 2014 ; Wildenhain et al, 2015 ; KalantarMotamedi et al, 2018 ), and single agent efficacies (Gayvert et al, 2017 ) and/or combinations (Menden et al, 2018 ) measured across multiple cell lines (for recent reviews of compound combination modeling and perspectives, see Bulusu et al, 2016 ; Weinstein et al, 2017 ; Tsigelny, 2018 ). A disadvantage to many of these approaches is that they often require experimental knowledge of underlying biological interactions between drugs and disease, or chemogenomic or phenotypic readouts (Jansen et al, 2009 ; Bansal et al, 2014 ; Wildenhain et al, 2015 ; Menden et al, 2018 ).…”

Raloxifene as Treatment for Various Types of Brain Injuries and Neurodegenerative Diseases: A Good Start

“…This may explain why raloxifene effects on development of Alzheimer’s disease may show the large variability that is seen [ 41 , 98 , 101 ]. In addition, to emphasize apparent anti-viral, as well as anti-bacterial properties of raloxifene, it has been found that it can also be used in treatment against malaria [ 111 ].…”