COV-Drug Circuit interaction server for Covid-19 Drug Repurposing
Kamal Rawal#1, Prashant Singh1, Robin Sinha1, Priya Kumari1, Swarsat Kaushik Nath1, Sukriti Sahai1, Ridhima1, Sweety Dattatraya Shinde1, Nikita Garg1, Preeti P.1, Trapti Sharma1
Amity Institute of Biotechnology, Amity University, Uttar Pradesh, India
#Corresponding Author
Email ID: kamal.rawal@gmail.com
Centre for Computational Biology and Bioinformatics, AIB
Amity University, Noida.
Keywords: Bioinformatics, drug repurposing, artificial intelligence, COVID-19, molecular targets
Supplementary Data Website: https://tinyurl.com/CoV-circuit
COV-DRUGX Software Pipeline: http://drugx.kamalrawal.in/drugx/
Abstract
The search for an effective and potent drug for the treatment of COVID-19 is still a big challenge. Various researchers across the globe have turned to drug repurposing pipelines to find the efficient drugs for COVID-19. There have been several strategies used for the drug repurposing. In this study we propose a Drug-Circuit Interaction based drug repositioning. Here, we present a web enabled server where the users can submit drugs in SMILE format and the server provides all the COVID-19 related genes associated with the given drug and all the pathway circuits associated with the identified genes. This server will help the user for efficient and cost-effective selection of the drugs that targets crucial COVID-19 associated pathways.
1. Introduction
COVID-19 is announced as a global pandemic in 2020. Due to limited medications, the emergent outbreak of COVID-19 prompted by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) keeps spreading [Acter T et al., 2020]. Despite the diminishing developments of COVID-19, there is no drug still approved to have significant effects with no side effect on the treatment for COVID-19 patients [ Kumar D et al., 2021] and identification of such drugs is still a challenge [Sharma T et al., 2021]. Understanding the molecular mechanisms that mediate SARS-CoV-2 infection is key for the rapid development of efficient preventive or therapeutic interventions against the COVID-19 [Rian K et al., 2021].
Apart from the popularly used drug repositioning techniques, mechanistic modeling of signal transduction circuits, with ML algorithms, is a recent approach for drug repurposing models for COVID-19 [Zhou et al., 2020]. Mechanistic models of pathways provide a natural bridge from variations at the scale of gene activity (transcription) to variations in phenotype (at the level of cells, tissues, or organisms). These models of human signaling pathways have been successfully used to uncover specific molecular mechanisms behind different diseases, to reveal modes of action of drugs, and to suggest personalized treatments [Çubuk C et al., 2020]. Analysis of functionalities within pathways has been underscoring the potential critical role of these molecular circuits in cellular control of the SARS-CoV-2 life cycle [Zhou et al., 2019]. Riva et al. (2020) has also defined that signalling may act as a critical host–pathogen interaction circuit in controlling viral infection [Riva L et al., 2020].
There have been hallmarks in signal transduction circuits defining its functionalities such as (1) host–virus interaction, (2) inflammatory response, (3) immune activity, (4) antiviral defense, (5) endocytosis, (6) replication, and (7) energetics [Loucera et al., 2020]. All these have been incorporated in our drug-circuit module as well. The module emphasizes on the study of drugs which have a key functional role in potential covid 19 associated pathways and their associated genes.
2. Implementation
The COV-Drug-Circuit server is written in python programming language to predict the associated interacting genes to our query drug. A list of 1,00,272 drugs and its associated genes was extracted from DGIdb (Supplementary Table 1). We have also downloaded a list of human genes and their associated pathways that plays a role in COVID-19 which was extracted from the research study conducted by Loucera et al., 2020 (Supplementary Table 2). Loucera et al developed a machine-learning algorithm that learns interactions between target proteins of drugs and their specific signaling circuits in the COVID-19 disease map. They have provided a list of 186 genes and 42 associated circuits. There are 299 total gene and circuit combinations that play a role in the pathogenesis of COVID-19. Along with this gene-circuit data, they have also provided the predictions on cell functionalities that trigger the COVID-19 symptoms or disease hallmarks. These hallmarks include host-virus interaction, inflammatory response, immune activity, anti-viral defense, endocytosis, replication and energetics.
We have utilized these 299 gene-circuit combinations in the COV-Drug-Circuit tool to identify the drug-gene, drug-circuit interaction and COVID-19 triggering cell functionalities.
3. Usage
The server provides the genes and circuits for a given drug name which is associated with COVID-19. Users provide the drug(s) name (seperated by pipe in a text file) or SMILE notation of the drugs (separated by newline character in a text file). The algorithm takes the provided drug names as a query and searches for its associated genes, and further looks for associated circuit information in the dataset. Then based on the circuit, it provides the functional information such as host-virus interaction, inflammatory response, immune activity, antiviral defense, endocytosis, replication and energetics. If the COV-Drug-Circuit server has found an associated circuit to our query drug in the dataset, then the drug will be scored 1 otherwise 0.
4. Result and Discussion
As a case study, we have collected three drug datasets i.e, 1,000 FDA approved drugs, 261 positive drugs and 37 drugs from machine learning study (Suvarna et al., 2021). The FDA approved drugs were extracted from the DrugBank database (https://go.drugbank.com/) run on the server (Supplementary Table 3). The intermediate result file has been analysed and found that duloxetine and atropine were found to have the highest number of circuit hits, i.e. 40 (Supplementary Table 4). For 652 drugs there were no circuits found. The distribution of the FDA approved drugs revealed that 46 drugs were found to be involved in above 10 pathways (Supplementary Figure 1).
The positive drugs dataset (261 drugs) collected from various literature were subjected to analysis on this tool and the intermediate file was obtained (Supplementary Table 5). Bortezomib was found to have the highest numbers of circuit hits (25 hits) (Supplementary Table 6). We have plotted the distribution of the total number of circuit hits against the total number of drugs (Supplementary Figure 2). 46 drugs were found to be involved in above 10 pathways. Total 214 drugs were found commonly related to approximately 5 circuits.
In another experiment, we have extracted 37 drugs from the study reported by Suvarna et al in the year 2021 (Suvarna et al., 2021). Suverna et al predicted 37 drugs as the prognostic markers for the COVID-19 using proteomics and machine learning approach. Those drugs were used as samples for our server and the predicted intermediate file is collected (Supplementary Table 7). The resultant file was analysed and top drugs were found including quercetin, ribavirin, valproic acid, rapamycin and verapamil which consist of maximum circuit hits (Supplementary Table 8). Further, we have plotted the distribution of the drugs with the common drugs found (Supplementary Figure 3).
The intermediate file of results consists of 23 columns. The “DRUG” column represents the drug name, the “VALUE” column gives the prediction (either 0 or 1), the “IN_DATABASE_VALUE” column describes the availability of the drug (ranges from 1 to -1). The “TOTAL_NUMBER_OF_CIRCUIT”, “NUMBER_OF_CIRCUIT_HIT”, and “NUMBER_OF_GENE_HIT” columns show the total number of circuit available in the database of the server, total number of circuit found from the database and total number of genes found from that particular circuit, respectively. The “Host-virus interaction”, “inflammatory response”, “Immune activity”, “Anti-viral defense”, “Endocytosis”, “Replication” and “Energetics” columns gives idea whether the given drug is involved in these or not. The “S1” column predicts the probability of the occurrence of the drug in the circuit. Rest all columns i.e., “CIRCUIT_HIT”, “GENE_HIT”, “Anti-viral defense HIT”, “Endocytosis HIT”, “Host-virus interaction HIT”, “Immune activity HIT”, “Inflammatory response HIT” and “Replication HIT” gives the name of hits found from the database.
The formula to calculate similarity is as follows:
5. Conclusion
Drug-circuit interaction prediction can play an important role in rational drug repositioning approaches. Information regarding various genes interacting with drugs and their associated functionalities can help better evaluate the functional role of query drugs.
In this work we utilize The Drug Gene Interaction Database DGIdb for the prediction of interacting genes and the type of interaction with query drugs which would further help predict the associated circuit and other functional related information with the gene, and hence the Query Drug. There are a number of ways to find potential interacting drugs with the COVID-covid 19 associated circuits. Here, we have combined the data from investigational study and system biology approach to yield a promising tool to better understand the biological response of the drug.
6.
References
Acter, T., Uddin, N., Das, J., Akhter, A., Choudhury, T. R., & Kim, S. (2020). Evolution of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) as coronavirus disease 2019 (COVID-19) pandemic: A global health emergency. Science of the Total Environment, 730, 138996.
Kumar, D., & Trivedi, N. (2021). Disease-drug and drug-drug interaction in COVID-19: risk and assessment. Biomedicine & Pharmacotherapy, 111642.
Sharma, T., Abohashrh, M., Baig, M. H., Dong, J. J., Alam, M. M., Ahmad, I., & Irfan, S. (2021). Screening of drug databank against WT and mutant main protease of SARS-CoV-2: Towards finding potential compound for repurposing against COVID-19. Saudi Journal of Biological Sciences, 28(5), 3152-3159.
Rian, K., Esteban-Medina, M., Hidalgo, M. R., Çubuk, C., Falco, M. M., Loucera, C., ... & Dopazo, J. (2021). Mechanistic modeling of the SARS-CoV-2 disease map. BioData Mining, 14(1), 1-8.
Zhou, Y., Wang, F., Tang, J., Nussinov, R., & Cheng, F. (2020). Artificial intelligence in COVID-19 drug repurposing. The Lancet Digital Health.
Çubuk, C., Can, F. E., Peña-Chilet, M., & Dopazo, J. (2020). Mechanistic Models of Signaling Pathways Reveal the Drug Action Mechanisms behind Gender-Specific Gene Expression for Cancer Treatments. Cells, 9(7), 1579.
Zhou, Y., Zhou, B., Pache, L., Chang, M., Khodabakhshi, A. H., Tanaseichuk, O., ... & Chanda, S. K. (2019). Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nature communications, 10(1), 1-10.
Riva, L., Yuan, S., Yin, X., Martin-Sancho, L., Matsunaga, N., Burgstaller-Muehlbacher, S., ... & Chanda, S. K. (2020). A large-scale drug repositioning survey for SARS-CoV-2 antivirals. BioRxiv.
Loucera, C., Esteban-Medina, M., Rian, K., Falco, M. M., Dopazo, J., & Peña-Chilet, M. (2020). Drug repurposing for COVID-19 using machine learning and mechanistic models of signal transduction circuits related to SARS-CoV-2 infection. Signal transduction and targeted therapy, 5(1), 1-3.