INTEGRATION OF MOLECULAR DOCKING AND MACHINE LEARNING FOR DRUG REPURPOSING IN THE DISCOVERY OF SARS-COV-2 ANTIVIRAL AGENTS
SARS-CoV-2 has emerged as a global pandemic, claiming approximately 6 million lives annually by the year 2023. First identified in December 2019 in Wuhan, China, the rapid spread of this virus has rendered it perilous, prompting researchers to fervently seek antiviral remedies. Urgent steps are r...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/79768 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | SARS-CoV-2 has emerged as a global pandemic, claiming approximately 6 million
lives annually by the year 2023. First identified in December 2019 in Wuhan, China,
the rapid spread of this virus has rendered it perilous, prompting researchers to
fervently seek antiviral remedies. Urgent steps are required in the discovery of new
drugs, leading to the adoption of drug repurposing through in silico methods to
expedite the process. This study employs molecular docking and machine learning
techniques to assess the binding affinity between drug molecules and the target
proteins of SARS-CoV-2. Molecular docking is executed using YASARA software,
while machine learning leverages the Deepchem library, specializing in drug
repurposing and design. The outcomes obtained through both methods manifest as
binding affinity, representing the strength of interaction between a molecule and the
active site of its target protein. Six target proteins (1s9p, 5r7y, 5rl6, 5s6x, 6m71, and
7rgq) are selected for this study, with a total of 82 antiviral drugs and 45 anticancer drugs employed. Molecular docking results provide binding affinity values
for each ligand-protein interaction, which are subsequently used to construct a
random forest model for predictions. Through molecular docking, Remdesivir is
identified as the drug with the highest binding affinity for protein target 1s9p,
Rimantadine for 5rl6 and 5s6x, and Oseltamivir for 5r7y, 6m71, and 7egq. The
machine learning model, utilizing the random forest algorithm, demonstrates
effective predictive capabilities with a mean absolute error (MAE) of 0.484365 and
an average ranking difference with molecular docking of 4.02. This suggests that the
machine learning model is proficient in predicting potential drug candidates,
providing valuable insights for drug repurposing against SARS-CoV-2 |
|---|