<p>WGCNA was used to identify DR-related PANoptosis genes, and the LASSO, SVM-RFE, and Random Forest machine learning models were then employed to identify key PANoptosis-related genes. The lncRNA-miRNA-TLR3 networks were constructed, and the levels of hub lncRNAs, hub miRNAs and TLR3 were measured in a high-glucose cell model. The luciferase reporter assay was employed to validate the interactions between Gm12610 and miR-758-3p, as well as between miR-758-3p and Tlr3. In the GSE102485 and GSE185011 cohort, TLR3 expression was significantly elevated in the DR samples compared to controls. Data from the GSE236627 cohort indicated a marked upregulation of Tlr3 in retinal microglia of db/db mice relative to normal controls. GSEA results showed that AGE-RAGE signaling pathway in diabetic complications and NF-kappa B signaling pathway were enriched and activated in the high-TLR3 expression (H-TLR3) group. Additionally, M2 macrophage infiltration was reduced in this group. Through the ENCORI and TargetScan databases, two ceRNA networks were constructed: C15orf54/CDKN2B-AS1-hsa-miR-758-3p-TLR3 and CDKN2B-AS1-hsa-miR-374b-5p-TLR3. In vitro experiments validated that high glucose-stimulated microglia showed significantly increased levels of Tlr3 and Gm12610 (mouse CDKN2B-AS1), but decreased levels of miR-374b-5p and miR-758-3p compared to the control group. Luciferase assays confirmed direct binding between Gm12610 and miR-758-3p, and between miR-758-3p and Tlr3. Our study identifies TLR3 as a key PANoptosis-related gene in DR, suggesting it potential as a therapeutic target for DR.</p>