Page 109 - MI-2-2

Microbes & Immunity Big data and DNN-based DTI model in CHP

potentially slow or prevent the progression of lung fibrosis. CHP treatment. Our methodology included system
This is crucial for CHP patients, as managing fibrosis is a order detection methods for systematic identification of
major treatment challenge. Studies suggest that masitinib GWGEN, which can effectively eliminate false-positive
may be beneficial in treating both fibrotic and non-fibrotic interactions and regulations within candidate GWGEN,
forms of HP by regulating immune responses and reducing resulting in the accurate system identification of the real
tissue damage. Compared to traditional broad-spectrum GWGEN for both CHP and non-CHP. Through big data
89
immunosuppressants, masitinib’s targeted molecular exploration, we constructed comprehensive genetic and
pathway action offers a more specific treatment approach. epigenetic biological networks of CHP and non-CHP.
Primaquine is traditionally used to treat and prevent Employing the PNP method, we extracted core GWGEN
malaria by disrupting the mitochondrial function of for the annotation of KEGG signaling pathways to obtain
Plasmodium. In addition, it has been found to have potential the core signaling pathways of CHP and non-CHP. By
in regulating the immune system, particularly by inhibiting comparing these core signal pathways of CHP and non-
T-cell overactivation and reducing inflammation, which is CHP, we investigated the pathogenetic mechanism of
significant in treating immune responses involved in HP. CHP and then identified key biomarkers crucial to CHP-
Primaquine also has antifibrotic effects, which can slow or induced cellular fibrosis. To discover candidate drugs and
90
prevent the progression of lung fibrosis in HP patients. their drug targets (biomarkers), we trained a DNN-based
It effectively downregulates AKT1 and NF-κB expression DTI model using the DTI database, predicting interaction
levels in the MAPK signaling pathway, which play crucial probabilities using the drug-target feature vectors. For drug
roles in the metabolic changes of fibrotic cells. Although targets, we applied stringent criteria – drug regulation
more clinical research and trials are needed to verify capacity, low toxicity, and high sensitivity – to screen
primaquine’s specific effects in HP treatment, its potential potential molecular drugs from the predicted candidate
as a molecular drug is promising, providing new ideas for molecular drugs. Consequently, we identified azathioprine,
developing more effective HP treatments. masitinib, and primaquine as an optimal multi-molecular
drug combination for CHP treatment, targeting critical
There are a few limitations to this study. First, it relies on biomarkers including AKT1, TNF, CCL-20, CDC23, and
GSE86618 microarray data, and the quality and completeness CXCL1. Our findings demonstrate that the integration of
of this data directly affect the accuracy of the results. If the multiple types of genome-wide genomic data of CHP and
dataset contains errors or incomplete data, it may lead to non-CHP with the systems biology method can significantly
biased conclusions. Second, the study’s conclusions are enhance our understanding of key biomarkers in the role of
mainly based on mathematical models and bioinformatics pathogenetics involved in CHP-induced cellular fibrosis.
analysis, lacking clinical experimental support. Predictions The proposed combination of the systems biology method
not validated by clinical and biological experiments may and systematic drug discovery design offers a promising
face challenges in practical drug applications, necessitating new direction for the treatment of chronic lung fibrosis
further clinical and biological experiments to ensure their progression. Future studies will benefit from incorporating
effectiveness and safety. In addition, although multiple more diverse genomic data types for epigenetic and
drug combinations were predicted, their actual effects and epigenomic regulation, which will further refine our
interactions need to be confirmed through in vivo and in understanding and improve therapeutic strategies for CHP.
vitro experiments. Finally, the model predictions may not
fully consider potential synergistic or antagonistic effects Acknowledgments
between different drugs. These limitations need to be None.
addressed in future research to improve the reliability and
applicability of drug repurposing strategies. Funding
5. Conclusion None.
In this study, we developed and validated a novel approach Conflict of interest
that combines systems biology to investigate the complex
pathogenetic mechanisms underlying CHP-induced The authors declare that they do not have competing
cellular fibrosis. Utilizing extensive database mining, interests.
we analyzed whole-genome data along with genetic and Author contributions
epigenetic networks to identify significant biomarkers of
pathogenetic mechanisms as drug targets from a systemic Conceptualization: Sung-Yu Lin, Bor-Sen Chen
drug design perspective for multi-molecular drugs of Formal analysis: Sung-Yu Lin, Bor-Sen Chen

Volume 2 Issue 2 (2025) 101 doi: 10.36922/mi.4620