International journal of chronic obstructive pulmonary disease · 2026
Revealing Biomarkers and Candidate Drugs for Chronic Obstructive Pulmonary Disease Using Bioinformatics and Deep Learning.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Wu Tianbiao, Wang Aiping, Xu Jinggang
Аннотация
BACKGROUND: There is a dearth of precise treatments capable of effectively decelerating the progression of chronic obstructive pulmonary disease (COPD). Consequently, there is an urgent necessity to discover novel potential biomarkers and rapidly translatable therapeutic drugs.
METHODS: The GSE57148, GSE38974, and GSE173896 datasets were included in this study. Cell clustering and annotation were conducted via Seurat, and disease-related genes were screened through an integrated approach combining Scissor-based prioritization, differential expression profiling, and weighted gene co-expression network analysis. Subsequently, 12 machine learning algorithms were used to construct models and identify biomarkers. The mechanisms of immune microenvironment remodeling were analyzed through cell communication and pseudo-time series analysis. Furthermore, an in vitro COPD model was constructed based on BEAS-2B cells and validated by RT-qPCR. Finally, a deep learning model and molecular docking were utilized to conduct AI-driven drug repurposing screening targeting biomarkers.
RESULTS: A total of 67 intersection genes were obtained. APRT, COX8A, EDF1 and RPL37 were identified as biomarkers for COPD. Further cell communication analysis revealed that CD4 naive T cells mediate immune microenvironment remodeling via the HLA-E-CD94/NKG2A inhibitory axis. Additionally, in vitro experiments confirmed that APRT, COX8A, and EDF1 were significantly underexpressed in the model group. Among these, EDF1 represented a previously unreported candidate biomarker in COPD and was therefore selected as a putative target for AI-driven drug repurposing. Deep learning and molecular docking targeting the novel target EDF1 identified several FDA - approved drugs with binding energies ≤ -5.0 kcal/mol. Among these drugs, doxycycline exhibited a stable binding conformation and strong binding activity with EDF1, indicating high potential for drug repurposing.
CONCLUSION: This study identified four potential biomarkers for COPD and screened several candidate drugs targeting EDF1, providing new targets and translational strategies for the early diagnosis and precision treatment of COPD. EDF1 is a novel candidate target not previously reported in COPD; however, functional validation is required before it can be considered a therapeutic target.
METHODS: The GSE57148, GSE38974, and GSE173896 datasets were included in this study. Cell clustering and annotation were conducted via Seurat, and disease-related genes were screened through an integrated approach combining Scissor-based prioritization, differential expression profiling, and weighted gene co-expression network analysis. Subsequently, 12 machine learning algorithms were used to construct models and identify biomarkers. The mechanisms of immune microenvironment remodeling were analyzed through cell communication and pseudo-time series analysis. Furthermore, an in vitro COPD model was constructed based on BEAS-2B cells and validated by RT-qPCR. Finally, a deep learning model and molecular docking were utilized to conduct AI-driven drug repurposing screening targeting biomarkers.
RESULTS: A total of 67 intersection genes were obtained. APRT, COX8A, EDF1 and RPL37 were identified as biomarkers for COPD. Further cell communication analysis revealed that CD4 naive T cells mediate immune microenvironment remodeling via the HLA-E-CD94/NKG2A inhibitory axis. Additionally, in vitro experiments confirmed that APRT, COX8A, and EDF1 were significantly underexpressed in the model group. Among these, EDF1 represented a previously unreported candidate biomarker in COPD and was therefore selected as a putative target for AI-driven drug repurposing. Deep learning and molecular docking targeting the novel target EDF1 identified several FDA - approved drugs with binding energies ≤ -5.0 kcal/mol. Among these drugs, doxycycline exhibited a stable binding conformation and strong binding activity with EDF1, indicating high potential for drug repurposing.
CONCLUSION: This study identified four potential biomarkers for COPD and screened several candidate drugs targeting EDF1, providing new targets and translational strategies for the early diagnosis and precision treatment of COPD. EDF1 is a novel candidate target not previously reported in COPD; however, functional validation is required before it can be considered a therapeutic target.