Integrated single-cell and bulk RNA-seq analysis reveals prognostic stemness genes in leiomyosarcoma

Aiju Lou¹Yu Cai²Tuquan Zheng²Long Zhang²Yupeng Hu²Huachi Li³Ying Lang^4*

• ¹Liwan Central Hospital of Guangzhou, Guangzhou, China
• ²Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, Guangdong Province, China
• ³Hubei Provincial Colorectal Cancer Clinical Medical Research Center, Wuhan, Hebei Province, China
• ⁴Yancheng Third People's Hospital, Yancheng, China

In this study, we used single-cell RNA-seq data of leiomyosarcoma from Gene Expression Omnibus (GEO), along with bulk RNA-seq and clinical data from The Cancer Genome Atlas Program (TCGA), to identify prognostic markers. By applying machine learning algorithms to the single-cell data, we identified malignant cells and calculated their stemness index. We found that cells with a high stemness index exhibited a more complex tumor immune microenvironment (TIME) and cell interactions compared to cells with a low stemness index. Differential gene analysis was performed to identify genes differentially expressed between high and low stemness index cells. To identify prognostic markers, we first applied univariate Cox regression to screen candidate genes, followed by Lasso regression for feature selection, and finally conducted multivariate Cox regression for model construction and survival analysis. This sequential approach led to the identification of six prognostic markers associated with tumor cell proliferation: BOP1, CTBP1, DSE, PMSD10, SRPK1, and HACD4. These genes could serve as potential therapeutic targets for personalized treatment of leiomyosarcoma (LMS).