Confirmatory Results

Repurposing of Anti-infectives for the Management of Onchocerciasis using Machine Learning and Protein Docking Studies

Cyril Tetteh, Bernice Ampomah, Mahmood B. Oppong, Michael Lartey, Paul Owusu Donkor, Kwabena F.M. Opuni, Lawrence A. Adutwum
doi: https://doi.org/10.1101/2023.02.26.530155

Abstract

The discovery of new drugs for the treatment of neglected tropical diseases (NTDs) is hampered by the lack of financial reward associated with their development. A combination of exploratory data analysis, machine learning (ML) and molecular docking studies were used to evaluate 58 anti-infective agents to identify those with the potential to be repurposed for the management of onchocerciasis, an NTD. Out of the 58 test drugs, 14 were predicted by at least five ML models to be useful in managing onchocerciasis. Molecular docking studies using glycine receptor subunit α-3, gamma-aminobutyric acid receptor subunit β-3 and glutamate-gated chloride channel of the 14 drugs showed binding affinities comparable to that of doramectin and pyrvinium which are known onchocerciasis drugs. Diminazine, trimetinib, triclabendazole, cridanimod, vandetinib and trametinib were the top agents showing high binding affinities from the molecular docking studies. The binding affinities of diminazine and cridanimod are similar to doramectin and pyrvinium which have demonstrated activity against onchocerciasis. The outcome of this study shows the potential of using these 14 drugs to manage onchocerciasis.

1. Introduction

Neglected tropical diseases (NTDs) are a group of diseases that affect low-income populations in the developing countries of Africa, Asia, and the Americas 1–3. Approximately one billion people, predominantly in developing countries living in remote rural areas, urban slums, or conflict zones, are at risk of NTDs 3,4. Onchocerciasis (river blindness or Robles’ disease) is a skin and eye infection caused by Onchocerca volvulus, a parasitic worm5,6. The disease is transmitted through the bites of infected blackflies, and symptoms of the disease include bumps under the skin, extreme itching, psoriasis, and loss of skin elasticity, among others. The most severe manifestations are ocular lesions that can progress to visual impairment and blindness7.

Ivermectin remains the main drug used to treat the disease8–10. Side effects such as neurotoxicity, central nervous system depression and liver toxicity have been reported11,12. Studies conducted in Ghana revealed that subjects repeatedly treated with ivermectin exhibited sub-optimal responses to the drug13. Even though several approaches are used to control the spread and the management of onchocerciasis, research on finding new therapeutic agents is not actively pursued14. Other drugs used in the past for managing onchocerciasis included diethylcarbamazine, suramin and moxidectin.15–24 Suramin use had to be discouraged due to the risk of optic atrophy16, while diethylcarbamazine is less effective even at high doses with frequent and severe side effects25. Pharmacokinetic studies on moxidectin show that it is an attractive long-acting therapeutic option for treating human onchocerciasis31 and was approved by the United States FDA in 201823,24.

New drug discovery is confronted with many challenges including huge expenses, time, and the increasing stringency of regulatory agencies in evaluating the efficacy of new drug candidates26–29. Research into finding new chemotherapeutic agents for NTDs is highly unattractive because endemic areas are largely poverty-stricken. Due to the onerous nature, associated risk, and low return on research and development investments, there is limited interest in developing medications for treating NTDs including onchocerciasis 28.

Drug repurposing/repositioning is the process of finding new therapeutic indications for existing drugs reducing research cost on toxicities and activity26,30–32. This reduces the time and cost of drug discovery as information on doses and toxicities is already established. Repurposing of already existing drugs implies that research and costs associated with all pre-clinical work are also exempted31,32. As an alternative, it offers scientists and pharmaceutical companies an efficient strategy to identify novel targets and uses for already approved pharmaceutical agents. This provides the advantage of reducing the effort, time, and cost of every drug development stage for an area that is plagued by a need for affordable and effective therapies32. In fact, the latest drug moxidectin is a veterinary anthelminthic which has been repurposed for to manage onchocerciasis. Even though high throughput screening allows for the screening of several thousands of agents in a relatively short time, techniques that can weed out agents with low probability of being hits will be beneficial.

Recently, the use of ML methods to screen compounds in databases to find new indications have become popular27,33–36. Among others, concepts such as artificial intelligence, deep learning, re-enforcement learning have been used33,34,37–40. Halicin, an effective antibiotic was discovered using deep learning approaches38. ML methods have been used to identify drug candidates, which have been repurposed for managing diseases such as Alzheimer’s, cancers, and COVID-1933,34,37,41–45. Unfortunately, there is a dearth of literature on the use of ML methods to identify potential drug candidates that can be repurposed for NTDs such as onchocerciasis.

In this work, exploratory data analysis (EDA) and ML techniques were used to screen several anti-infective agents to identify potential candidates that may have activity against onchocerciasis. Subsequently, molecular docking studies were done on potential candidates to determine their binding interactions and binding affinities using targets that typically bind to drugs used for onchocerciasis.

2. Experimental Section

2.1. Dataset generation and Pre-processing

Three-dimensional (3D) structures of 602 anti-infective drugs were obtained from online repositories including DrugBank46, PubChem47, and ChemSpider48. Where the 3D conformers of the compounds were unavailable, their two-dimensional (2D) conformers were converted to their optimized 3D structures using WebMO (version 21.0)49. The web-based cheminformatics platform, online chemical modeling environment (OCHEM) and AlvaDesc® (version 2.0.4) were used to generate molecular descriptors 50–52. This yielded a dataset matrix of 602 rows × 5,666 columns (rows × columns).

Descriptors with entries for less than 30% of the data were eliminated. This reduced the number of descriptors to 4,474. Of the 602 samples, 16 are/were used (or have demonstrated activity) in the management of onchocerciasis while 588 were not. The data thus comprised of two classes, which are onchocerciasis and non-onchocerciasis drugs which were labeled oncho+ and oncho-, respectively. The data was mean centered and auto scaled to unit variance.

2.2. Data Analysis

2.2.1. Linear Discriminant Analysis (LDA) and Cluster Analysis (k-Means)

Using the scaled data, LDA was performed as a dimensionality reduction step using singular value decomposition solver. The significant uses of the drugs (12 indications) were used as the target. This was followed by cluster analysis using the k-means cluster algorithm to identify drugs like the oncho+. The pairwise distance between the oncho+ drugs and all other anti-infectives were determined. Five (5) anti-infectives closest to each onchocerciasis drug were selected and put into a test set. In total, 56 samples were captured in this group and were thus put into a test set. This reduced the training set data to 546 (i.e., 16 oncho+, 530 oncho-, 56 test drugs). Doramectin and pyrvinium, agents demonstrating activity against onchocerciasis were added to the test drugs to evaluate our proposed routine. Hence, the distribution of the data set for subsequent analysis became 14 oncho+, 530 oncho-, and 58 test drugs (including doramectin and pyrvinium).

2.2.2. Class Imbalance and Machine Learning

The imbalance in the data, (14 oncho+, 530 oncho-) was corrected using the Synthetic Minority Oversampling Technique53,54 (SMOTE). A total of 516 artificial compounds were generated to balance the data. Thus, the final dataset for subsequent ML modeling was 1060 (530 oncho+ & 530 oncho-)

The dataset was split into 2/3 (707 drugs) and 1/3 (353 drugs) for training and validation sets, respectively. Ten supervised learning algorithms namely, Support Vector Machines (SVM), Random Forest (RF), Logistic Regression (LR), k-Nearest Neighbor (KNN) classification, Stochastic Gradient Boosting (SGB), Boosted Decision Trees (BDT), Naïve Bayes Classifiers (NBC), AdaBoost (ABC), Multi-Layer Perceptron (MLP), Gaussian Naïve Bayes (GNB) were used. These ML models were generated with the training set data using scikit-learn. Hyperparameter tuning was performed automatically to achieve the optimum model. The models’ performance was evaluated using their prediction accuracy as defined by Long et.al.55 on the external validation set. The models with prediction accuracies of at least 80% were used to evaluate the 58 test compounds.

2.2.3. Protein Target Retrieval and Virtual Screening (Molecular Docking)

The binding affinity of test drugs predicted to have oncho+ activities were evaluated using molecular docking studies. Four proteins identified as relevant to the management of onchocerciasis were used as targets. These were glycine receptor subunit α-3 (GRS-α-3, UniProt ID: O75311)56, gamma-aminobutyric acid receptor subunit β-3 (GABA-β-3, UniProt ID: P28472)57, Glutamate-gated chloride channel (GluCl, PDB ID: Q25634)58, and prostaglandin G/H synthase 1 (PGHS, UniProt ID: P23219)59. The PDB files for the crystal structures of these four proteins were downloaded from RCSB Database (https://www.rcsb.org/) and the SWISS-MODEL Repository (https://swissmodel.expasy.org/). The retrieved structures were cleaned up with Discovery Studio version 21.1.0 (BIOVIA, San Diego) 60 to correct issues associated with incomplete structures due to missing atoms or water, the presence of multimers and interaction partners of the receptor molecule.

Virtual screening was performed using the Python Prescription Virtual Screening tool (PyRx 0.8) 61 containing AutoDock Vina module 62. The prepared protein structure was fed into the PyRx tool along with the 3D conformers of the test drugs predicted to be oncho+. The drug and protein molecules were converted to a PDBQT file using the AutoDock module of the PyRx tool. A grid box was then constructed to define docking spaces. The dimensions of the grid box were set along the x, y, and z axes at: 104.2, 102.5, 137.9Å for GABA-β-3, 130.9, 87.2, 126.6 Å for GRS-α-3, 154.1, 75.0, 122.7 Å for GluCl and 81.8, 69.9, 75.3 Å for PGHS. These parameters were set to encompass the entire 3D structure of each protein so that the ligand could freely move and rotate in the docking space. The 2D and 3D interactions between the protein-ligand were analyzed using Discovery Studio version 21.1.0 (BIOVIA, San Diego) 60.

3. Results and Discussion

The discovery of new drugs can be a long, tedious, and costly endeavour. Consequently, pharmaceutical companies invest in projects with higher chances of financial success. Thus, research into the discovery of some medications may be lacking even if they are needed by the population. The case is much worse for NTDs as the people in need of these drugs lack the financial capabilities to afford them. Drug repurposing presents a relatively cheaper route to discovering new drugs by finding new indications for already existing drugs. In this study, we explored EDA and ML techniques to identify anti-infectives that may have the potential to be repurposed for the management of onchocerciasis. To support the ML predictions, molecular docking studies were performed using known targets of drugs used to manage onchocerciasis to determine the interactions and binding affinities of the predicted agents.

ML algorithms benefit from the availability of large data sets. Unfortunately, fewer agents are used in the management of onchocerciasis leading to an imbalance in the dataset. With only 14 samples assigned to the oncho+ class, the high number of descriptors relative to the sample size could lead to overfitting. Using a typical feature selection algorithm would not be ideal, hence, LDA was performed as a dimensionality reduction step.

Figure 1 (a and b) shows the LDA results using the major indications of the anti-infectives (Figure S1) as the output. The LDA latent variables (LDs) scatter plot for the first four LDs are shown in Figure S2 in the SI. The LDs in LDA are constructed to maximize the separation between the samples, i.e., significant uses. Subsequently, 11 LDs were determined to be ideal since they capture 99.77 % of the variability in the data.

Figure 1.
Figure 1.

Visualizing the LDA model of anti-infective agents using the major indications/drugs (a) LD1 vs LD2 and b) LD3 vs LD2).

To find agents that could be repurposed for managing onchocerciasis from the data, an unsupervised learning approach, k-means cluster analysis was employed. In cluster analysis, drug samples that have similar properties/indications are expected to be grouped closer to each other. The k-means clustering algorithm was used due to its fast computation and general ability to produce tighter clusters63,64. Using the Folk Mallows and Rand Scores, an optimal k value of 11 was identified. The 11 clusters were just one less than the total number of indications of the drugs in the data set. Subsequently, the 14 oncho+ samples were identified and the pairwise distance from these 14 drugs and the other samples were computed. The five (5) drugs that clustered closest to the oncho+ drugs were selected. It is expected that their closer proximity to the oncho+ drugs implied the likelihood to have similar therapeutic properties. A total of 70 drugs would be found in the test set, however, there were 56 drugs as some agents were found to be close to more than one oncho+ drug. Figure S3 in SI is LD2 vs LD3 plot of the LDA model showing the location of oncho+(red), oncho- (blue) and test drugs (orange). It must be highlighted here that doramectin and pyrvinium with known oncho+ were added to investigate whether this approach would have identified them as potentially useful drugs65,66.

Generation of ML models with imbalanced data is not ideal as it leads to high model accuracies even if the models perform woefully in the minority class. The Imbalanced-learn SMOTE algorithm was implemented to generate samples to balance the data in the minority class. In total, 516 artificial oncho+ samples were generated to balance the dataset. A plot of LD1 vs LD2 and LD2 vs LD3 for the dataset with and without the artificial samples can be found in Figure S4 in the SI.

ML models were generated using the balanced data as described earlier. Two-thirds of the data was used for the training set whiles a third was used for external validation. The following ML algorithms were used: SVM, RF, GNB, NB, KNN, LR, SGB, BDT, MLP and ADB. All these algorithms were used since one could not tell which would perform better for this dataset. Hyperparameter tuning was performed automatically. The model prediction accuracies on the external validation sets were used to evaluate the ML models. Figure 2 shows the accuracy of the ten different machine learning algorithms in predicting anti-infective agents as, specifically when tested on an external validation set of data. It can be seen here that with the same dataset and pre-processing, different algorithms give different prediction accuracies. Hence, valuable information may be lost if a single model is relied on. Consequently, the test drugs in this study were evaluated using an ensemble of the models. However, since all the models have different prediction accuracies, a threshold of at least 80% was selected. Seven (7) models met this criterion: SVM, RF, KNN, BDT, SGB, MLP and ADB. Subsequently, the seven (7) models were used to predict the onchocercidal activity of the 58 test drugs. The threshold prediction probability was set to a default value of 0.5. Thus, for a model, any drug with a y predicted ≥ 0.5 is deemed to have the potential to have onchocercidal activity. The number of drugs predicted by each algorithm is shown in Figure S5 in the SI. Figure S5 shows KNN predicted 38 of the test drugs as having the potential to be used for onchocerciasis with SGB predicting the least 13. Figure 3a-g shows the y predicted psrobabilities of the 58 test drugs. The results indicate varying predicted probabilities for the test drugs by different ML algorithms. These variations in predicted probabilities for different algorithms are expected due to the various approximation in the objective functions used in solving ML problems. Thus, using several algorithms allows us to truly identify which drugs have true potential. Due to the variation in the number of predicted samples by each ML model, drugs selected by at least 5 of the 7 ML models were selected for molecular docking studies.

Figure 2.
Figure 2.

The model prediction accuracies of the ten (10) machine learning algorithms using the external validation set data.

Figure 3.
Figure 3.

The y-predicted probability plots of the 16 test samples (including pyrvinium and doramectin) for the top seven best performing ML algorithms based on the model prediction accuracy on the external validation set (a) SVM, b) RF, c) KNN, d) BDT, e) SGB, f) MLP and g) ADB. The red dashed line represents the 0.5 predicted probability value above which a drug is classified as oncho+.

Table S1 in SI shows the list of the test drugs and their predictions by each of the 7 selected models. Sixteen (16) drugs representing 27.59 % of the test data were predicted to be oncho+. It is noteworthy that onchocercidal drugs doramectin and pyrivinium were predicted as oncho+ by all 7 models. This affirms that a combination of EDA and ML truly has the potential to aid in the repurposing of anti-infectives for NTDs.

Molecular Docking Studies

The drug of choice in the management of onchocerciasis is ivermectin. Its antiparasitic activity is attributed to its binding to ivermectin-sensitive glutamate-gated chloride channel (GluCl) and glycine receptor subunit α-3 (GRS-α-3)67–69. In mouse models, evidence shows that the anticonvulsant effect of Ivermectin is due to its binding to gamma-aminobutyric acid receptor subunit β-3 (GABA-β-3)69. Diethylcarbamazine, another anti onchocerciasis agent is thought to be involved in sensitizing the micro-filariae to phagocytosis through its interaction with Prostaglandin G/H synthase 1 (PGHS) in the arachidonic acid metabolic pathway70. Thus, for any of the predicted drugs to have some activity against onchocerciasis it must bind to some, if not all, of these target proteins. Thus, the binding interaction with these targets can provide useful information to further boost the confidence in the drugs. The binding affinity, Kd measures the interaction between two molecules - a macromolecule (e.g., protein//target) and a ligand (drug) 71. Low Kd indicates higher affinity whiles higher Kd indicates lower affinity.

The binding interaction between these four receptors and the 16 drugs was performed as described earlier. The results of these experiments are shown in Figure 4. It is evident from Figure 4 that all the drugs predicted by the ML algorithms have binding interactions with all the protein targets studied. It is not surprising that our evaluation drugs, doramectin and pyrvinium are the 1st and 2nd drugs in all targets except GRS-α-3 where pyrvinium was 3rd. The detailed binding interactions of the top three for each target excluding the evaluating drugs were examined.

Figure 4.
Figure 4.

The binding affinity (Kd) of the 16 test samples (including pyrvinium and doramectin) for the GluCl, GRS-a-3, GABA-b-3 and PGHS receptors.

The binding interaction of the top three drugs for GRS-α-3 namely, vandetinib, trimetinib, and diminazine are show in Figure 5. Vandetanib showed a binding affinity of -9.5 kcal/mol to GRS-α-3 which is higher than pyrvinium, the evaluation drug. Vandetanib forms a hydrogen bond and an anion-pi interaction with the ALA26 and ASP86 residues while Trametinib forms pi-sigma bond and a halogen interaction with the ALA254 and PRO250 residues. Diminazine forms four hydrogen bonds with SER82, TYR161, THR113 and ARG27 as well as a pi-pi T-shaped interaction with PRO10 residues of GRS-α-3.

Figure 5.
Figure 5.

The binding interaction between GRS-a-3 and Diminazine, Trametinib and Vandetinib.

It is interesting to note that vandetanib, diminazine and trametinib also turned to be the top three drugs that bind to GABA-β-3. From Figure 6, it can be seen that vandetinib forms three hydrogen bonds with ASP48, VAL50 and VAL53, a pi-sulphur bond with MET137 and a halogen interaction with GLU182. Vandetanib also forms a pi-alkyl bond with LYS274, a residue found to link with pyrivinium through pi-donor interaction. Diminazine forms a pi-pi stacked interaction with TYR62, a pi-sigma interaction with ALA201 and several hydrogen bonds with ASN41, TYR97, SER156, TYR157, ALA174 and LYS173. Trametinib forms a pi-pi stacked interaction with PHE306, a pi-pi T-shaped interaction with TYR304 and two hydrogen bonds with ARG428 and TYR304.

Figure 6.
Figure 6.

The binding interaction between GABA-β-3and Diminazine, Trametinib and Vandetinib

The binding affinities for cridanimod, diminazine and vandetanib with GluCl were -7.8, -7.2 and -7.1 kcal/mol respectively. Cridanimod is a small molecule functionally related to acridone and shown to have antineoplastic adjuvant activity. The compound interacts with GluCl via hydrogen bonding at GLU25, ASN45, ASN46, and GLN103; pi-anion at GLU25 and pi-alkyl at VAL40 and ARG41. Diminazine forms two hydrogen bonds at PRO216 and ILE70, pi-pi stacked at TYR254 similar to pyrivinium. However, there was shown to be an unfavorable donor-donor interaction with SER217. Vandetanib forms a halogen bond with GLN258 and pi-cation interaction with LYS305 (Figure 7).

Figure 7.
Figure 7.

The binding interaction between GluCl and Diminazine, Trametinib and Cridanimod

In the case of PGHS, diminazine, trimetinib and triclabendazole demonstrated the highest binding. Detailed binding interactions are shown in Figure 8. Triclabendazole is a narrow-spectrum anthelminthic with activity against Fasciola and Paragonimus spp. It shows a binding affinity of -8.2 kcal/mol and interacted with PGHS via hydrogen bonding with the amino acid residues CYS41 and CYS47 as well as halogen bonding and pi-alkyl interaction with CYS47 and PRO153 respectively. Trametinib forms one hydrogen bond with ARG120, a pi-sigma bond with VAL119 and a pi-cation interaction with ARG83 while Diminazine forms numerous hydrogen bonds with the residues TRP387, PHE210, MET391, ALA199, THR206, TYR385 and THR212 which was also found to interact with Doramectin through hydrogen bonding. Diminazine also forms three pi-interactions with PHE210, HIS388 and ALA202 and a pi-cation interaction with HIS386.

Figure 8.
Figure 8.

The binding interaction between PGHS and Triclabendazole, Trametinib and Diminazine.

4. Conclusion

Using exploratory data analysis, machine learning and molecular docking studies, 14 anti-infectives have been identified to possess the potential to be repurposed for the management of onchocerciasis. Out of these 14 drugs, six showed strong binding interactions comparable to drugs used in the management of onchocerciasis. These six are diminazine, trimetinib, triclabendazole, cridanimod, vandetinib and trametinib. Of these drugs, diminazine exhibits the highest potential to be used for onchocerciasis. These compounds could also serve as lead compounds in the discovery of new and improved therapies for onchocerciasis.

The techniques and routines implemented here can be applied to other research areas to find new indications for existing drugs.

5. Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

1 Data Availability Statement

The datasets for this study can be found in the https://github.com/lawrenceadutwum/ntdrepurpose.

6. References

  1. 1.↵
    1. Hotez PJ,
    2. Aksoy S,
    3. Brindley PJ,
    4. Kamhawi S.
    What constitutes a neglected tropical disease? PLoS Negl Trop Dis. 2020;14(1). doi:10.1371/JOURNAL.PNTD.0008001
    OpenUrlCrossRef
  2. 2.
    1. Ocholaid EA,
    2. Karanja DMS,
    3. Elliott SJ
    . The impact of neglected tropical diseases (Ntds) on health and wellbeing in sub-saharan africa (ssa): A case study of Kenya. PLoS Negl Trop Dis. 2021;15(2). doi:10.1371/journal.pntd.0009131
    OpenUrlCrossRef
  3. 3.↵
    1. Hopkins AD
    . Neglected tropical diseases in Africa: A new paradigm. Int Health. 2015;8. doi:10.1093/inthealth/ihv077
    OpenUrlCrossRef
  4. 4.↵
    1. Blair JMA,
    2. Richmond GE,
    3. Bailey AM,
    4. Ivens A,
    5. Piddock LJ v
    . Choice of bacterial growth medium alters the transcriptome and phenotype of Salmonella enterica Serovar Typhimurium. PLoS One. 2013;8(5):e63912.
    OpenUrlCrossRefPubMed
  5. 5.↵
    1. Hendy A,
    2. Krüger A,
    3. Pfarr K, et al
    . The blackfly vectors and transmission of Onchocerca volvulus in Mahenge, south eastern Tanzania. Acta Trop. 2018;181. doi:10.1016/j.actatropica.2018.01.009
    OpenUrlCrossRef
  6. 6.↵
    1. Hotterbeekx A,
    2. Namale Ssonko V,
    3. Oyet W,
    4. Lakwo T,
    5. Idro R.
    Neurological manifestations in Onchocerca volvulus infection: A review. Brain Res Bull. 2019;145. doi:10.1016/j.brainresbull.2018.08.024
    OpenUrlCrossRef
  7. 7.↵
    1. Winthrop KL,
    2. Furtado JM,
    3. Silva JC,
    4. Resnikoff S,
    5. Lansingh VC
    . River blindness: An old disease on the brink of elimination and control. J Glob Infect Dis. 2011;3(2). doi:10.4103/0974-777X.81692
    OpenUrlCrossRef
  8. 8.↵
    1. Eneanya OA,
    2. Koudou BG,
    3. Aboulaye M, et al
    . Progress towards onchocerciasis elimination in côte d’ivoire: A geospatial modelling study. PLoS Negl Trop Dis. 2021;15(2). doi:10.1371/journal.pntd.0009091
    OpenUrlCrossRef
  9. 9.
    1. Babalola OE,
    2. Bassi A.
    Impact assessment study after 27 years of community-directed treatment with ivermectin in Galadimawa, Kaduna State, Nigeria. Niger Postgrad Med J. 2017;24(1). doi:10.4103/npmj.npmj_6_17
    OpenUrlCrossRef
  10. 10.↵
    1. de Smet E,
    2. Metanmo S,
    3. Mbelesso P, et al
    . Focus of ongoing onchocerciasis transmission close to bangui, central african republic. Pathogens. 2020;9(5). doi:10.3390/pathogens9050337
    OpenUrlCrossRef
  11. 11.↵
    1. Somo RM,
    2. Ngosso A,
    3. Dinga JS,
    4. Enyong PA,
    5. Fobi G.
    A community-based trial of ivermectin for onchocerciasis control in the forest of southwestern Cameroon: Clinical and parasitologic findings after three treatments. In: American Journal of Tropical Medicine and Hygiene. Vol 48. ; 1993. doi:10.4269/ajtmh.1993.48.9
    OpenUrlCrossRef
  12. 12.↵
    1. Veit O,
    2. Beck B,
    3. Steuerwald M,
    4. Hatz C.
    First case of ivermectin-induced severe hepatitis. Trans R Soc Trop Med Hyg. 2006;100(8). doi:10.1016/j.trstmh.2006.02.003
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Osei-Atweneboana MY,
    2. Awadzi K,
    3. Attah SK,
    4. Boakye DA,
    5. Gyapong JO,
    6. Prichard RK
    . Phenotypic evidence of emerging ivermectin resistance in Onchocerca volvulus. PLoS Negl Trop Dis. 2011;5(3). doi:10.1371/journal.pntd.0000998
    OpenUrlCrossRef
  14. 14.↵
    1. Boussinesq M,
    2. Fobi G,
    3. Kuesel AC
    . Alternative treatment strategies to accelerate the elimination of onchocerciasis. Int Health. 2018;10. doi:10.1093/inthealth/ihx054
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Hawking F.
    Suramin: With Special Reference to Onchocerciasis. Adv Pharmacol. 1978;15(C). doi:10.1016/S1054-3589(08)60486-X
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Thylefors B,
    2. Rolland A.
    The risk of optic atrophy following suramin treatment of ocular onchocerciasis. Bull World Health Organ. 1979;57(3).
  17. 17.
    1. Anderson J,
    2. Fuglsang H.
    Further studies on the treatment of ocular onchocerciasis with diethylcarbamazine and suramin. British Journal of Ophthalmology. 1978;62(7). doi:10.1136/bjo.62.7.450
    OpenUrlFREE Full Text
  18. 18.
    1. Turner HC,
    2. Walker M,
    3. Attah SK, et al
    . The potential impact of moxidectin on onchocerciasis elimination in Africa: An economic evaluation based on the Phase II clinical trial data. Parasit Vectors. 2015;8(1). doi:10.1186/s13071-015-0779-4
    OpenUrlCrossRefPubMed
  19. 19.
    1. Milton P,
    2. Hamley JID,
    3. Walker M,
    4. Basáñez MG.
    Moxidectin: an oral treatment for human onchocerciasis. Expert Rev Anti Infect Ther. Published online 2020. doi:10.1080/14787210.2020.1792772
    OpenUrlCrossRef
  20. 20.
    1. Milton P,
    2. Hamley JID,
    3. Walker M,
    4. Basáñez G,
    5. Basáñez MG.
    Expert Review of Anti-infective Therapy Moxidectin: an oral treatment for human onchocerciasis Moxidectin: an oral treatment for human onchocerciasis. Expert Rev Anti Infect Ther. 2020;18(11).
  21. 21.
    1. Keller L,
    2. Palmeirim MS,
    3. Ame SM, et al
    . Efficacy and Safety of Ascending Dosages of Moxidectin and Moxidectin-albendazole Against Trichuris trichiura in Adolescents: A Randomized Controlled Trial. Clinical Infectious Diseases. 2020;70(6). doi:10.1093/cid/ciz326
    OpenUrlCrossRef
  22. 22.
    1. Tan B,
    2. Opoku N,
    3. Attah SK, et al
    . Pharmacokinetics of oral moxidectin in individuals with Onchocerca volvulus infection. PLoS Negl Trop Dis. 2022;16(3). doi:10.1371/journal.pntd.0010005
    OpenUrlCrossRef
  23. 23.↵
    1. Opoku NO,
    2. Bakajika DK,
    3. Kanza EM, et al
    . Single dose moxidectin versus ivermectin for Onchocerca volvulus infection in Ghana, Liberia, and the Democratic Republic of the Congo: a randomised, controlled, double-blind phase 3 trial. The Lancet. 2018;392(10154). doi:10.1016/S0140-6736(17)32844-1
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Awadzi K,
    2. Opoku NO,
    3. Attah SK,
    4. Lazdins-Helds J,
    5. Kuesel AC
    . A Randomized, Single-Ascending-Dose, Ivermectin-Controlled, Double-Blind Study of Moxidectin in Onchocerca volvulus Infection. PLoS Negl Trop Dis. 2014;8(6). doi:10.1371/journal.pntd.0002953
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Albiez EJ,
    2. Newland HS,
    3. White AT, et al
    . Chemotherapy of onchocerciasis with high doses of diethylcarbamazine or a single dose of ivermectin: Microfilaria levels and side effects. Tropical Medicine and Parasitology. 1988;39(1).
  26. 26.↵
    1. Wall G,
    2. Lopez-Ribot JL
    . Screening repurposing libraries for identification of drugs with novel antifungal activity. Antimicrob Agents Chemother. 2020;64(9). doi:10.1128/AAC.00924-20
    OpenUrlAbstract/FREE Full Text
  27. 27.↵
    1. Méndez-Lucio O,
    2. Baillif B,
    3. Clevert DA,
    4. Rouquié D,
    5. Wichard J.
    De novo generation of hit-like molecules from gene expression signatures using artificial intelligence. Nat Commun. 2020;11(1). doi:10.1038/s41467-019-13807-w
    OpenUrlCrossRef
  28. 28.↵
    1. Cherla A,
    2. Renwick M,
    3. Jha A,
    4. Mossialos E.
    Cost-effectiveness of cancer drugs: Comparative analysis of the United States and England. EClinicalMedicine. 2020;29-30. doi:10.1016/j.eclinm.2020.100625
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Mohs RC,
    2. Greig NH
    . Drug discovery and development: Role of basic biological research. Alzheimer’s and Dementia: Translational Research and Clinical Interventions. 2017;3(4). doi:10.1016/j.trci.2017.10.005
    OpenUrlCrossRef
  30. 30.↵
    1. Park K.
    A review of computational drug repurposing. Transl Clin Pharmacol. 2019;27(2). doi:10.12793/tcp.2019.27.2.59
    OpenUrlCrossRef
  31. 31.↵
    1. Fatima S,
    2. Bhaskar A,
    3. Dwivedi VP
    . Repurposing Immunomodulatory Drugs to Combat Tuberculosis. Front Immunol. 2021;12. doi:10.3389/fimmu.2021.645485
    OpenUrlCrossRef
  32. 32.↵
    1. Sleigh SH,
    2. Barton CL
    . Repurposing strategies for therapeutics. Pharmaceut Med. 2010;24(3). doi:10.1007/BF03256811
    OpenUrlCrossRef
  33. 33.↵
    1. Issa NT,
    2. Stathias V,
    3. Schürer S,
    4. Dakshanamurthy S.
    Machine and deep learning approaches for cancer drug repurposing. Semin Cancer Biol. 2021;68. doi:10.1016/j.semcancer.2019.12.011
    OpenUrlCrossRef
  34. 34.↵
    1. Tanoli Z,
    2. Vähä-Koskela M,
    3. Aittokallio T.
    Artificial intelligence, machine learning, and drug repurposing in cancer. Expert Opin Drug Discov. 2021;16(9). doi:10.1080/17460441.2021.1883585
    OpenUrlCrossRef
  35. 35.
    1. Nam Y,
    2. Kim M,
    3. Chang HS,
    4. Shin H.
    Drug repurposing with network reinforcement. BMC Bioinformatics. 2019;20. doi:10.1186/s12859-019-2858-6
    OpenUrlCrossRef
  36. 36.↵
    1. Masoudi-Sobhanzadeh Y,
    2. Omidi Y,
    3. Amanlou M,
    4. Masoudi-Nejad A.
    Drug databases and their contributions to drug repurposing. Genomics. 2020;112(2). doi:10.1016/j.ygeno.2019.06.021
    OpenUrlCrossRef
  37. 37.↵
    1. Mohanty S,
    2. Harun AI
    3. Rashid M,
    4. Mridul M,
    5. Mohanty C,
    6. Swayamsiddha S.
    Application of Artificial Intelligence in COVID-19 drug repurposing. Diabetes and Metabolic Syndrome: Clinical Research and Reviews. 2020;14(5). doi:10.1016/j.dsx.2020.06.068
    OpenUrlCrossRefPubMed
  38. 38.↵
    1. Stokes JM,
    2. Yang K,
    3. Swanson K, et al
    . A Deep Learning Approach to Antibiotic Discovery. Cell. 2020;180(4). doi:10.1016/j.cell.2020.01.021
    OpenUrlCrossRefPubMed
  39. 39.
    1. Rodriguez S,
    2. Hug C,
    3. Todorov P, et al
    . Machine learning identifies candidates for drug repurposing in Alzheimer’s disease. Nat Commun. 2021;12(1). doi:10.1038/s41467-021-21330-0
    OpenUrlCrossRef
  40. 40.↵
    1. Anderson E,
    2. Havener TM,
    3. Zorn KM, et al
    . Synergistic drug combinations and machine learning for drug repurposing in chordoma. Sci Rep. 2020;10(1). doi:10.1038/s41598-020-70026-w
    OpenUrlCrossRef
  41. 41.↵
    1. Aghdam R,
    2. Habibi M,
    3. Taheri G.
    Using informative features in machine learning based method for COVID-19 drug repurposing. J Cheminform. 2021;13(1). doi:10.1186/s13321-021-00553-9
    OpenUrlCrossRef
  42. 42.
    1. Mohapatra S,
    2. Nath P,
    3. Chatterjee M, et al
    . Repurposing therapeutics for COVID-19: Rapid prediction of commercially available drugs through machine learning and docking. PLoS One. 2020;15(11 November). doi:10.1371/journal.pone.0241543
    OpenUrlCrossRef
  43. 43.
    1. Zame WR,
    2. Bica I,
    3. Shen C, et al
    . Machine learning for clinical trials in the era of COVID-19. Stat Biopharm Res. 2020;12(4). doi:10.1080/19466315.2020.1797867
    OpenUrlCrossRef
  44. 44.
    1. Peiffer-Smadja N,
    2. Maatoug R,
    3. Lescure FX,
    4. D’Ortenzio E,
    5. Pineau J,
    6. King JR
    . Machine Learning for COVID-19 needs global collaboration and data-sharing. Nat Mach Intell. 2020;2(6). doi:10.1038/s42256-020-0181-6
    OpenUrlCrossRef
  45. 45.↵
    1. Rodriguez S,
    2. Hug C,
    3. Todorov P, et al
    . Machine learning identifies novel candidates for drug repurposing in Alzheimer’s Disease. bioRxiv. Published online 2020. doi:10.1101/2020.05.15.098749
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    1. Wishart DS,
    2. Knox C,
    3. Guo AC, et al
    . DrugBank: a comprehensive resource for in silico drug discovery and exploration. Nucleic Acids Res. 2006;34(Database issue):668–672. doi:10.1093/nar/gkj067
    OpenUrlCrossRef
  47. 47.↵
    1. Kim S,
    2. Chen J,
    3. Cheng T, et al
    . PubChem in 2021: New data content and improved web interfaces. Nucleic Acids Res. 2021;49(D1):D1388–D1395. doi:10.1093/nar/gkaa971
    OpenUrlCrossRefPubMed
  48. 48.↵
    1. A. Chemist
    . ChemSpider SyntheticPages. 2001. http://cssp.chemspider.com/123
  49. 49.↵
    1. Polik WF,
    2. Schmidt JR
    . WebMO: Web-based computational chemistry calculations in education and research. Wiley Interdiscip Rev Comput Mol Sci. 2022;12(1). doi:10.1002/wcms.1554
    OpenUrlCrossRef
  50. 50.↵
    1. Vorberg S,
    2. Tetko I v.
    Modeling the biodegradability of chemical compounds using the online chemical modeling environment (OCHEM). Mol Inform. 2014;33(1). doi:10.1002/minf.201300030
    OpenUrlCrossRef
  51. 51.
    1. Sushko I,
    2. Novotarskyi S,
    3. Körner R, et al
    . Online chemical modeling environment (OCHEM): Web platform for data storage, model development and publishing of chemical information. J Comput Aided Mol Des. 2011;25(6). doi:10.1007/s10822-011-9440-2
    OpenUrlCrossRefPubMed
  52. 52.↵
    1. Mauri A.
    AlvaDesc: A tool to calculate and analyze molecular descriptors and fingerprints. In: Methods in Pharmacology and Toxicology. ; 2020. doi:10.1007/978-1-0716-0150-1_32
    OpenUrlCrossRef
  53. 53.↵
    1. Hussein AS,
    2. Li T,
    3. Yohannese CW,
    4. Bashir K.
    A-SMOTE: A new preprocessing approach for highly imbalanced datasets by improving SMOTE. International Journal of Computational Intelligence Systems. 2019;12(2). doi:10.2991/ijcis.d.191114.002
    OpenUrlCrossRef
  54. 54.↵
    1. Fernandez A,
    2. Garcia S,
    3. Herrera F,
    4. Chawla N v.
    SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary. Journal of Artificial Intelligence Research. 2018;61:863–905. doi:10.1613/jair.1.11192
    OpenUrlCrossRef
  55. 55.↵
    1. Loong T wey.
    Clinical review Understanding sensitivity and specificity with the right. Bmj. 2003;327(September):716–719. doi:10.1136/bmj.327.7417.716
    OpenUrlFREE Full Text
  56. 56.↵
    1. Huang X,
    2. Chen H,
    3. Michelsen K,
    4. Schneider S,
    5. Shaffer PL
    . Crystal structure of human glycine receptor-α3 bound to antagonist strychnine. Nature. 2015;526(7572). doi:10.1038/nature14972
    OpenUrlCrossRefPubMed
  57. 57.↵
    1. Miller PS,
    2. Aricescu AR
    . Crystal structure of a human GABAA receptor. Nature. 2014;512(7514). doi:10.1038/nature13293
    OpenUrlCrossRefPubMed
  58. 58.↵
    1. Hibbs RE,
    2. Gouaux E.
    Principles of activation and permeation in an anion-selective Cys-loop receptor. Nature. 2011;474(7349). doi:10.1038/nature10139
    OpenUrlCrossRef
  59. 59.↵
    1. Miciaccia M,
    2. Belviso BD,
    3. Iaselli M, et al
    . Three-dimensional structure of human cyclooxygenase (hCOX)-1. Sci Rep. 2021;11(1). doi:10.1038/s41598-021-83438-z
    OpenUrlCrossRef
  60. 60.↵
    1. Dassault Systèmes
    . BIOVIA. http://www3ds.com/products-services/biovia/. Published online 2021.
  61. 61.↵
    1. Dallakyan S,
    2. Olson AJ
    . Small-Molecule Library Screening by Docking with PyRx. In: Methods in Molecular Biology. Springer New York; 2015:243–250. doi:10.1007/978-1-4939-2269-7_19
    OpenUrlCrossRefPubMed
  62. 62.↵
    1. Trott O,
    2. Olson AJ
    . Software News and Update AutoDock Vina: Improving the Speed and Accuracy of Docking with a New Scoring Function, Efficient Optimization, and Multithreading. J Comput Chem. 2010;31:455–461. doi:10.1002/jcc.21334
    OpenUrlCrossRefPubMedWeb of Science
  63. 63.↵
    1. von Luxburg U.
    A tutorial on spectral clustering. Stat Comput. 2007;17(4). doi:10.1007/s11222-007-9033-z
    OpenUrlCrossRefWeb of Science
  64. 64.↵
    1. Tavallali P,
    2. Tavallali P,
    3. Singhal M.
    K-means tree: an optimal clustering tree for unsupervised learning. Journal of Supercomputing. 2021;77(5). doi:10.1007/s11227-020-03436-2
    OpenUrlCrossRef
  65. 65.↵
    1. Gunderson EL,
    2. Bryant C,
    3. Bulman CA, et al
    . Pyrvinium Pamoate and Structural Analogs Are Early Macrofilaricide Leads. Pharmaceuticals. 2022;15(2). doi:10.3390/ph15020189
    OpenUrlCrossRef
  66. 66.↵
    1. Tagboto S,
    2. Orish V.
    Drug development for onchocerciasis-the past, the present and the future. Frontiers in Tropical Diseases. 2022;3(July). doi:10.3389/fitd.2022.953061
    OpenUrlCrossRef
  67. 67.↵
    1. Cully DF,
    2. Vassilatis DK,
    3. Liu KK, et al
    . Cloning of an avermectin-sensitive glutamate-gated chloride channel from Caenorhabditis elegans. Nature. 1994;371(6499). doi:10.1038/371707a0
    OpenUrlCrossRef
  68. 68.
    1. Dk V,
    2. Jp A,
    3. Rha P, et al
    . Genetic and biochemical evidence for a novel avermectin-sensitive chloride channel in Caenorhabditis elegans - Isolation and characterization. Journal of Biological Chemistry. 1997;272.
  69. 69.↵
    1. Shan Q,
    2. Haddrill JL,
    3. Lynch JW
    . Ivermectin, an Unconventional Agonist of the Glycine Receptor Chloride Channel. Journal of Biological Chemistry. 2001;276(16). doi:10.1074/jbc.M011264200
    OpenUrlAbstract/FREE Full Text
  70. 70.↵
    1. McGarry HF,
    2. Plant LD,
    3. Taylor MJ
    . Diethylcarbamazine activity against Brugia malayi microfilariae is dependent on inducible nitric-oxide synthase and the cyclooxygenase pathway. Filaria J. 2005;4. doi:10.1186/1475-2883-4-4
    OpenUrlCrossRef
  71. 71.↵
    1. Pantsar T,
    2. Poso A.
    Binding affinity via docking: Fact and fiction. Molecules. 2018;23(8). doi:10.3390/molecules23081899
    OpenUrlCrossRef
Back to top
Posted February 27, 2023.
Download PDF
Email
Repurposing of Anti-infectives for the Management of Onchocerciasis using Machine Learning and Protein Docking Studies
Cyril Tetteh, Bernice Ampomah, Mahmood B. Oppong, Michael Lartey, Paul Owusu Donkor, Kwabena F.M. Opuni, Lawrence A. Adutwum
bioRxiv 2023.02.26.530155; doi: https://doi.org/10.1101/2023.02.26.530155
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area