Open Source Bayesian Models. 1. Application to ADME/Tox and Drug Discovery Datasets

Clark, Alex M.; Dole, Krishna; Coulon-Spektor, Anna; McNutt, Andrew; Grass, George M.; Freundlich, Joel S.; Reynolds, Robert C.; Ekins, Sean

doi:10.1021/acs.jcim.5b00143

Cited by 101 publications

(174 citation statements)

References 148 publications

Supporting

Mentioning

172

Contrasting

Order By: Relevance

“…It is therefore imperative that factors that determine the disposition of a pharmaceutical compound in an organism, such as absorption, distribution, metabolism, excretion, and toxicity (ADMET), are assessed early enough in the drug development process so that any potential issues can be addressed. While many of these properties can be reliably predicted (40,41), there is still tremendous value to generating in vitro data for drug candidates. To that end, kinetic solubility (42), CYP inhibition (43), metabolic stability (44), Caco-2 permeability (45), and plasma protein binding (in mice and humans) (46) all were evaluated.…”

Section: Resultsmentioning

confidence: 99%

Efficacy of Tilorone Dihydrochloride against Ebola Virus Infection

Ekins

Lingerfelt

Comer

et al. 2018

Antimicrob Agents Chemother

Self Cite

View full text Add to dashboard Cite

Tilorone dihydrochloride (tilorone) is a small-molecule, orally bioavailable drug that is used clinically as an antiviral outside the United States. A machinelearning model trained on anti-Ebola virus (EBOV) screening data previously identified tilorone as a potent in vitro EBOV inhibitor, making it a candidate for the treatment of Ebola virus disease (EVD). In the present study, a series of in vitro ADMET (absorption, distribution, metabolism, excretion, toxicity) assays demonstrated the drug has excellent solubility, high Caco-2 permeability, was not a P-glycoprotein substrate, and had no inhibitory activity against five human CYP450 enzymes (3A4, 2D6, 2C19, 2C9, and 1A2). Tilorone was shown to have 52% human plasma protein binding with excellent plasma stability and a mouse liver microsome half-life of 48 min. Dose range-finding studies in mice demonstrated a maximum tolerated single dose of 100 mg/kg of body weight. A pharmacokinetics study in mice at 2-and 10-mg/kg dose levels showed that the drug is rapidly absorbed, has dosedependent increases in maximum concentration of unbound drug in plasma and areas under the concentration-time curve, and has a half-life of approximately 18 h in both males and females, although the exposure was ϳ2.5-fold higher in male mice. Tilorone doses of 25 and 50 mg/kg proved efficacious in protecting 90% of mice from a lethal challenge with mouse-adapted with once-daily intraperitoneal (i.p.) dosing for 8 days. A subsequent study showed that 30 mg/kg/day of tilorone given i.p. starting 2 or 24 h postchallenge and continuing through day 7 postinfection was fully protective, indicating promising activity for the treatment of EVD.

show abstract

Section: Resultsmentioning

confidence: 99%

Efficacy of Tilorone Dihydrochloride against Ebola Virus Infection

Ekins

Lingerfelt

Comer

et al. 2018

Antimicrob Agents Chemother

Self Cite

View full text Add to dashboard Cite

show abstract

“…We recently made “function class fingerprints of maximum diameter 6” (FCFP6) and “extended connectivity (ECFP6) fingerprints,” open source and have described their implementation with the Chemistry Development Kit (CDK) 50 components 41 . In addition we described an open source Bayesian algorithm that can be used with these descriptors 39, 40 . One way to make such models more accessible is to use mobile devices for their delivery and we have developed cheminformatics mobile apps 41, 51– 55 .…”

Section: Discussionmentioning

confidence: 99%

Machine learning models identify molecules active against the Ebola virus in vitro

et al. 2016

Self Cite

View full text Add to dashboard Cite

The search for small molecule inhibitors of Ebola virus (EBOV) has led to several high throughput screens over the past 3 years. These have identified a range of FDA-approved active pharmaceutical ingredients (APIs) with anti-EBOV activity in vitro and several of which are also active in a mouse infection model. There are millions of additional commercially-available molecules that could be screened for potential activities as anti-EBOV compounds. One way to prioritize compounds for testing is to generate computational models based on the high throughput screening data and then virtually screen compound libraries. In the current study, we have generated Bayesian machine learning models with viral pseudotype entry assay and the EBOV replication assay data. We have validated the models internally and externally. We have also used these models to computationally score the MicroSource library of drugs to select those likely to be potential inhibitors. Three of the highest scoring molecules that were not in the model training sets, quinacrine, pyronaridine and tilorone, were tested in vitro and had EC 50 values of 350, 420 and 230 nM, respectively. Pyronaridine is a component of a combination therapy for malaria that was recently approved by the European Medicines Agency, which may make it more readily accessible for clinical testing. Like other known antimalarial drugs active against EBOV, it shares the 4-aminoquinoline scaffold. Tilorone, is an investigational antiviral agent that has shown a broad array of biological activities including cell growth inhibition in cancer cells, antifibrotic properties, α7 nicotinic receptor agonist activity, radioprotective activity and activation of hypoxia inducible factor-1. Quinacrine is an antimalarial but also has use as an anthelmintic. Our results suggest data sets with less than 1,000 molecules can produce validated machine learning models that can in turn be utilized to identify novel EBOV inhibitors in vitro.

show abstract

“…Over the last decade we and others 15–17, 58–61 have increasingly focused on Bayesian approaches because of their ease of use and general applicability 56, 62–78 using molecular function class fingerprints 79, 80 of maximum diameter 6 and several other simple descriptors 81, 82 . Much of this work was centered on models for Mycobacterium tuberculosis 83–85 taking account of cytotoxicity and prospectively evaluating them to show high hit rates compared to random screening 85–87 .…”

Section: Introductionmentioning

confidence: 99%

“…Much of this work was centered on models for Mycobacterium tuberculosis 83–85 taking account of cytotoxicity and prospectively evaluating them to show high hit rates compared to random screening 85–87 . We have since followed this with datasets for Chagas disease 88 and Ebola 89 to repurpose approved drugs as well as model ADME properties such as aqueous solubility, mouse liver microsomal stability 90 , Caco-2 cell permeability 62 , toxicology datasets 91 and transporters 66, 92–97 . By making the fingerprints 98 , and Bayesian model building algorithm open source 21, 62 there is the potential to further expand on this work.…”

Section: Introductionmentioning

confidence: 99%

Comparison of Deep Learning With Multiple Machine Learning Methods and Metrics Using Diverse Drug Discovery Data Sets

Korotcov¹,

Tkachenko²,

et al. 2017

Self Cite

View full text Add to dashboard Cite

Machine learning methods have been applied to many datasets in pharmaceutical research for several decades. The relative ease and availability of fingerprint type molecular descriptors paired with Bayesian methods resulted in the widespread use of this approach for a diverse array of endpoints relevant to drug discovery. Deep learning is the latest machine learning algorithm attracting attention for many of pharmaceutical applications from docking to virtual screening. Deep learning is based on an artificial neural network with multiple hidden layers and has found considerable traction for many artificial intelligence applications. We have previously suggested the need for a comparison of different machine learning methods with deep learning across an array of varying datasets that is applicable to pharmaceutical research. Endpoints relevant to pharmaceutical research include absorption, distribution, metabolism, excretion and toxicity (ADME/Tox) properties, as well as activity against pathogens and drug discovery datasets. In this study, we have used datasets for solubility, probe-likeness, hERG, KCNQ1, bubonic plague, Chagas, tuberculosis and malaria to compare different machine learning methods using FCFP6 fingerprints. These datasets represent whole cell screens, individual proteins, physicochemical properties as well as a dataset with a complex endpoint. Our aim was to assess whether deep learning offered any improvement in testing when assessed using an array of metrics including AUC, F1 score, Cohen’s kappa, Matthews correlation coefficient and others. Based on ranked normalized scores for the metrics or datasets Deep Neural Networks (DNN) ranked higher than SVM, which in turn was ranked higher than all the other machine learning methods. Visualizing these properties for training and test sets using radar type plots indicates when models are inferior or perhaps over trained. These results also suggest the need for assessing deep learning further using multiple metrics with much larger scale comparisons, prospective testing as well as assessment of different fingerprints and DNN architectures beyond those used.

show abstract

Open Source Bayesian Models. 1. Application to ADME/Tox and Drug Discovery Datasets

Cited by 101 publications

References 148 publications

Efficacy of Tilorone Dihydrochloride against Ebola Virus Infection

Efficacy of Tilorone Dihydrochloride against Ebola Virus Infection

Machine learning models identify molecules active against the Ebola virus in vitro

Comparison of Deep Learning With Multiple Machine Learning Methods and Metrics Using Diverse Drug Discovery Data Sets

Contact Info

Product

Resources

About