Despite the well-established role of long non-coding RNAs (lncRNAs) across various biological processes, their mechanisms in acute myocardial infarction (AMI) are not fully elucidated. The GSE34198 dataset from the Gene Expression Omnibus (GEO) database, which comprised 49 specimens from individuals with AMI and 47 specimens from controls, was extracted and analysed using the weighted gene co-expression network analysis (WGCNA) package. Twenty-seven key genes were identified through a combination of the degree and gene significance (GS) values, and the
CDC42
(degree = 64),
JAK2
(degree = 41), and
CHUK
(degree = 30) genes were identified as having the top three-degree values among the 27 genes. Potential interactions between lncRNA, miRNAs and mRNAs were predicted using the starBase V3.0 database, and a lncRNA-miRNA-mRNA triple network containing the lncRNA
XIST
, twenty-one miRNAs and three hub genes (
CDC42
,
JAK2
and
CHUK
) was identified. RT–qPCR validation showed that the expression of the
JAK2
and
CDC42
genes and the lncRNA
XIST
was noticeably increased in samples from patients with AMI compared to normal samples. Pearson’s correlation analysis also proved that
JAK2
and
CDC42
expression levels correlated positively with lncRNA
XIST
expression levels. The area under ROC curve (AUC) of lncRNA
XIST
was 0.886, and the diagnostic efficacy of the lncRNA
XIST
was significantly better than that of
JAK2
and
CDC42
. The results suggested that the lncRNA
XIST
appears to be a risk factor for AMI likely through its ability to regulate
JAK2
and
CDC42
gene expressions, and it is expected to be a novel and reliable biomarker for the diagnosis of AMI.