IntroductionPeriodontitis is a widespread chronic inflammatory disorder that affects the periodontal supporting tissues, resulting in alveolar bone resorption, periodontal ligament destruction, tooth mobility, and eventual tooth loss. It represents a major cause of chewing dysfunction and impaired oral health-related quality of life1. Conventional diagnosis of periodontitis mainly depends on clinical indicators such as probing depth, clinical attachment loss, and bleeding on probing. These approaches are unable to detect early-stage molecular changes, leading to delayed intervention and poor prognosis in many patients1. Therefore, identifying sensitive and specific molecular biomarkers for early diagnosis is urgently needed to improve the management of periodontitis.Mitochondria are core organelles governing energy metabolism, redox balance, and cell survival. Mitochondrial dysfunction and excessive mitochondrial reactive oxygen species (mtROS) accumulation have been identified as crucial pathological events in periodontitis2,3. Increased mtROS promotes inflammatory activation, aggravates periodontal soft tissue damage, and stimulates osteoclast differentiation, thereby accelerating alveolar bone loss3. As a selective autophagic process that degrades damaged mitochondria, mitophagy serves as a central mechanism of mitochondrial quality control4. Mounting studies have confirmed that mitophagy is suppressed in periodontal tissues, which further disrupts mitochondrial homeostasis, amplifies oxidative stress, and promotes the progression of periodontitis5. Consequently, mitophagy-related genes (MRGs) are closely implicated in the onset and development of periodontitis and may serve as promising diagnostic biomarkers.Mitophagy also participates in bone homeostasis by regulating osteoclast activation and osteoblast apoptosis6,7,8. It acts as a protective response against oxidative damage in bone marrow mesenchymal stem cells and promotes osteogenic differentiation, indicating a potential role in periodontal regeneration7.Bioinformatics analysis based on public transcriptomic databases has become an efficient strategy for rapid screening of disease-associated biomarkers and exploration of pathogenic mechanisms9,10,11. Although several bioinformatics studies have been conducted in periodontitis, most focus on general inflammation or immune-related genes, while studies specifically targeting mitophagy-related biomarkers remain scarce. In addition, few studies have combined multi-algorithm bioinformatics screening with experimental verification using clinical samples, leading to insufficient reliability of the identified candidates.In the present study, we obtained periodontitis transcriptomic datasets from the Gene Expression Omnibus (GEO) database12 and performed differential expression analysis, WGCNA13, PPI analysis14,15, and functional enrichment analysis16,17,18,as well as multiple machine learning algorithms19,20,21 to identify mitophagy-related candidate biomarkers. The expression patterns and diagnostic value of the candidate genes were further validated in clinical gingival tissues by quantitative real-time polymerase chain reaction (RT-qPCR). We hypothesized that specific MRGs are abnormally expressed in periodontitis and can serve as potential biomarkers for early detection. This study aims to provide novel molecular targets for the early diagnosis and mechanistic investigation of periodontitis.Materials and methodsThe basic process of this study is shown below.Data source and data processingTranscriptomic microarray data of gingival tissues from periodontitis patients and healthy controls were downloaded from the Gene Expression Omnibus (GEO) database12. The training set GSE16134 contained 239 periodontitis samples and 69 control samples based on the GPL570 platform. The validation set GSE10334 contained 183 periodontitis samples and 64 control samples. Probe IDs were converted into gene symbols according to the platform annotation files. Gene expression data were normalized and analyzed using the “limma” package in R software (version 4.3.3). A total of 29 mitophagy-related genes (MRGs) were retrieved from the Pathway Unification database.Clinical sample collectionHuman gingival tissues were collected from patients who received treatment at the Affiliated Stomatological Hospital of Dalian Medical University between September 2024 and November 2024 (Table 1). Periodontitis tissues were obtained during periodontal flap surgery, gingivectomy, or tooth extraction. Healthy gingival tissues were collected from systemically healthy individuals during minor oral surgery. All samples were rinsed with saline, snap-frozen in liquid nitrogen, and stored at − 80 °C until use. This study was approved by the Ethics Committee of the Affiliated Stomatological Hospital of Dalian Medical University (No. 2024005). Written informed consent was obtained from all participants.Table 1 Basic information of clinical research subjects.Full size tableInclusion and exclusion criteriaPeriodontitis group(1)Age 30–75 years.(2)At least two teeth with clinical attachment loss (CAL) ≥ 3 mm and probing depth (PD) > 3 mm.(3)Positive bleeding on probing (BOP +).Healthy control group(1)Age 30–70 years.(2)PD ≤ 3 mm, no CAL, and BOP negative.(3)No systemic diseases or medications affecting periodontal status.Exclusion criteria(1)Smoking history.(2)Systemic diseases, immunodeficiency, or pregnancy.(3)Antimicrobial drugs used in the past 3 months.(4)Periodontal therapy in the past 6 months.Identification of differentially expressed genesDifferentially expressed genes (DEGs) between periodontitis and control samples in the training set GSE16134 were screened using the “limma” package in R. Thresholds were set as |log₂FC|> 0.5 and FDR