Objective To investigate the possible pathogenesis and clinical feature of mitochondrial encephalomyopathy with lactic acidemia and stroke-like episodes (MELAS) on the molecular level by bioinformatics analysis of differential expression genes. Methods The microarray information of the wild-type cell lines and the mutant cell line with high expression of mtDNA A3243G locus was downloaded from the GEO database, and the differential expression genes were obtained by the R platform. Then Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed, and paired comparison of gene microarray analysis to screen out the same differently expressed genes, and the protein-protein interaction (PPI) network was used to find the relation of key genes between MELAS and clinical symptoms. Results A total of 563 differential expression genes were obtained, of which 250 genes were up-regulated and 313 genes were down-regulated. GO enrichment analysis showed that the differential expression genes were mainly involved in the extracellular matrix (ECM) binding biological process, and KEGG pathway involved in phosphatidylinositol 3-kinase (PI3K)-serine/threonine kinase (AKT) signaling pathway, tranforming growth factor-β (TGF-β) signaling pathway, ECM receptor interaction pathway. Using the STRING platform, 9 hubs genes including GNG2, SDC2, ANXA1, FN1, TNC, CYR61, IGFBP3, LTBP1, SERPIND1 had been found. PAMR1, GLRX5, SNCA and other common differential genes were obtained by paired comparison of 3 groups of gene microarrays. Conclusions According to the above results, the PI3K-AKT signaling pathway, TGF-β signaling pathway, ECM receptor interaction pathway and the biological processes of ECM binding were involved in the process of the mitochondrial energy metabolism, which might be related to the pathogenesis of MELAS. GNG2, TNC, LTBP1, PAMR1, GLRX5, SNCA and other hubs genes might be related to the clinical manifestations of MELAS.