Integrating the analysis of transcriptome-wide association study and gene expression profile to identify genes associated with spondyloarthritis

Background: Spondyloarthritis(SpA) is a group of multi-factorial bone diseases influenced by genetic factors, environment and lifestyles. However, the genetic and pathogenic mechanism of SpA is still elusive. Methods Firstly, the tissue-specific transcriptome-wide association study (TWAS) of SpA was performed by utilizing the genome-wide association study (GWAS, including 3966 SpA patients and 452264 controls) summary data and gene expression weights of the whole blood and skeletal muscle. Secondly, the SpA-associated genes identified by TWAS were further compared with the differentially expressed genes(DEGs) detected by gene expression profile of SpA acquired from the Gene Expression Omnibus database (GEO, accession number:GSE58667). Finally, FUMA and Metascape tools were used to conduct gene functional enrichment and annotation analysis. Results TWAS detected 28 significant genes associated with SpA both in the whole blood and skeletal muscle, such as CTNNAL1 (P SM =0.0304, P WB =0.0096). Further comparing with gene expression profile of SpA, we identified 20 candidate genes which overlapped in TWAS, such as MCM4 (P TWAS =0.0132, P DEG =0.0275), KIAA1109 (P TWAS =0.0371,P DEG =0.0467). The enrichment analysis of the genes identified by TWAS detected 93 significant GO terms 33 and KEGG pathways, such as mitochondrion organization (GO:0007005, log 10 (P)= -4.29) and axon guidance(hsa04360, log 10 (P)= -4.26). Conclusion We identified multiple candidate genes genetically related to SpA. Our study may provide some novel clues for the further study of the genetic mechanism, diagnosis and treatment of SpA.