Screening of hub genes and immunocytes related to tendon injury based on bioinformatics and machine learning models

Tendon injury is a common and challenging clinical problem, and its healing process involves complex cellular and biological factors. Therefore, this study aims to reveal the mechanism of tendon healing and provide theoretical basis for clinical treatment. We first selected GSE26051 dataset from the GEO database and used R language to obtain 721 DEGs (459 up-regulated and 262 down-regulated). Subsequently, the 7378 genes of tendon injury obtained from the GeneCards database were intersected with DEGs to obtain 228 common genes. We constructed a PPI network of common genes using the STRING database, visualized it using Cytoscape software, and selected the top 10 (MYH6, MYL3, MYH1, MYH8, MYL1, TTN, TCAP, PKP2, ACTN2, CSRP3) genes through the CytoHubba plugin. We further identified hub genes (MYH1, MYH6, PKP2, MYH8) via machine learning models. Afterwards, the cytoskeleton in muscle cells and IL-17 signaling pathways were obtained by GO and KEGG analysis of common genes. Finally, the macrophages M2 was screened through immune infiltration analysis. This study revealed that hub genes such as MYH1, MYH6, MYH8 and PKP2 were mainly enriched in the cytoskeleton in muscle cells signaling pathway, and macrophage M2 played an important role in the inflammatory phase of tendon healing.