Objective To screen differentially expressed genes (DEG) related to brain metastasis in triple negative breast cancer（TNBC） using bioinformatics analysis, and to explore the potential mechanism affecting the prognosis of TNBC patients. Methods Datasets including GSE76250 (TNBC tissues and normal breast tissues) and GSE125989 (TNBC brain metastases tissue and TNBC primary lesion tissue) were retrieved from Gene Expression Omnibus (GEO) database, and DEG was screened using Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to identify potential related genes contributing to brain metastasis in TNBC. The association between gene expression levels and breast cancer prognosis was verified by clinical tissue samples from The Cancer Genome Atlas (TCGA) database, and gene enrichment analysis was performed using the gene set of KEGG to assess the signaling pathways that involved in genes related to brain metastasis. Results A total of 52 DEGs were screened in GSE125989 and GSE76250 datasets. Protein-protein interaction network (PPI) suggested that RAD51AP1 was an important gene related to brain metastasis in TNBC. Analysis of TCGA data showed that RAD51AP1 was significantly overexpressed in breast cancer tissues vs paracancerous tissues, and for different molecular subtypes of breast cancer,basal-like breast cancer had a higher level of RAD51AP1 than that in paracancerous tissues. Patients with breast cancer were divided into high expression and low expression of RAD51AP1 group according to the median expression level of RAD51AP1 in cancer tissues [log₂ (TPM+1)=3.85]. Survival analysis showed that the median survival of patients with high expression was lower than that of patients with low expression [median OS 3 873 days vs 3 945 days, P<0.05, HR=1.40 (1.01-1.94)]. GO, KEGG and gene set enrichment analysis (GSEA) showed that signaling pathways, such as cell cycle, DNA replication and mismatch repair, were significantly enriched in RAD51AP1 highly expressed phenotypes. PPI was constructed based on cell cycle and DNA damage repair signaling pathway related protein information, and proteins that directly interacted with RAD51AP1 were screened.It revealed proliferating cell nuclear antigen (PCNA) was strongly correlated with RAD51AP1. Conclusions RAD51AP1 is highly expressed in TNBC and TNBC brain metastasis tissues, which may serve as a potential biomarker for diagnosis of TNBC and prediction of poor prognosis, and mechanism on RAD51AP1 mediating brain metastasis may be related to PCNA.