Application of Single-Cell Sequencing and Machine Learning in Prognosis and Immune Profiling of Lung Adenocarcinoma: Exploring Disease Mechanisms and Treatment Strategies Based on Circadian Rhythm Gene Signatures.

: The circadian rhythm regulates important functions in the body, such as metabolism, the cell cycle, DNA repair, and immune balance. Disruption of this rhythm can contribute to the development of cancer. Circadian rhythm genes (CRGs) are attracting attention for their connection to various cancers. However, their roles in LUAD are not yet well understood. Additionally, our knowledge of how they function at both the bulk tissue and single-cell levels is limited. This gap hinders a complete understanding of how CRGs impact the development and outcomes of LUAD. : We selected 554 CRGs from public databases. We then obtained transcriptome data from TCGA and GEO. A total of 101 machine learning algorithm combinations were tested using 10 algorithms and 10-fold cross-validation. The best-performing model was based on Stepwise Cox regression and SuperPC. This model was validated with additional datasets. We also examined the relationships between CRGs, immune features, tumor mutation burden (TMB), and the response to immunotherapy. Drug sensitivity was also assessed. Single-cell data identified the cell types with active CRGs. Next, we performed qRT-PCR and other basic experiments to validate the expression of ARNTL2 in LUAD tissues and cell lines. The results indicated that ARNTL2 may play a key role in lung adenocarcinoma. : The CRG-based model clearly distinguished LUAD patients based on their risk. High-risk patients exhibited low immune activity, high TMB, and poor predicted responses to immunotherapy. Single-cell data revealed strong CRG signals in epithelial and fibroblast cells. These cell groups also displayed different communication patterns. Laboratory experiments showed that ARNTL2 was highly expressed in LUAD. It promoted cell growth, movement, and invasion. This suggests that ARNTL2 may play a role in promoting cancer. : This study developed a machine learning model based on CRGs. It can predict survival and immune status in LUAD patients. The research also identified ARNTL2 as a key gene that may contribute to cancer progression. These findings highlight the significance of the circadian rhythm in LUAD and provide new perspectives for diagnosis and treatment.