By Powerpuff Girls (Carmina Dinulescu, Diana Groza, Cosmina Sas)
The search for biomedical relations between chemical compounds (drugs, molecules) and protein targets is an important part of drug discovery.
Through the validation and prediction of drug-drug interactions (DDIs), researchers can improve the therapeutic efficacy of various drugs. Unfortunately, these interactions can also lead to the development of potentially harmful side effects. Early detection of potential issues with drug-drug interactions can help prevent the development of potentially harmful drugs.
Although in vitro experiments are widely used for the prediction of biochemical interactions, their reliability and cost-effectiveness are still limited by their complexity and time-consuming nature. In silico approaches have received more attention due to their increasing accuracy and cost-effectiveness.
Because drug structures can be naturally represented as graphs (with nodes and edges denoting chemical atoms and bonds, respectively), and protein structures can also be represented as a logical graph (with nodes and edges denoting amino acids and biochemical interactions, respectively), we can use graphical models to predict drug outcomes.
“In this work, we propose DeepDrug, a novel end-to-end deep learning framework for DDI and DTI predictions. DeepDrug takes in both drug SMILES strings and protein PDB (Protein Data Bank) inputs to characterize biochemical entities into graphical representations and utilizes GCNs to learn latent feature representations that give superior level of accuracy for predictive modeling. The competitive edge of graph-based architecture allows DeepDrug to incorporate both DDI and DTI predictions into a general framework. It also empowers DeepDrug to be applied to novel entities whose graphical representations can be extracted. Overall, through extensive experiments on existing DDI and DTI datasets and detailed comparison with other published methods, we demonstrate the promising performance of DeepDrug in drug-related interaction prediction tasks.”
The results of this research suggest that DeepDrug can be used not just to predict drug relationships, but also to uncover drug interaction processes. DeepDrug has demonstrated its effectiveness in a variety of DDI and DTI prediction tasks, although there is still potential for development.
Bibliography
Niciun comentariu:
Trimiteți un comentariu