Purpose of reviewTo summarize recent studies stratifying SLE patients into subgroups based on gene expression profiling and suggest future improvements for employing transcriptomic data to foster precision medicine.Recent findingsBioinformatic & machine learning pipelines have been employed to dissect the transcriptomic heterogeneity of lupus patients and identify more homogenous subgroups. Some examples include the use of unsupervised random forest and k-means clustering to separate adult SLE patients into seven clusters and hierarchical clustering of single-cell RNA-sequencing (scRNA-seq) of immune cells yielding four clusters in a cohort of adult SLE and pediatric SLE participants. Random forest classification of bulk RNA-seq data from sorted blood cells enabled prediction of high or low disease activity in European and Asian SLE patients. Inferred transcription factor activity stratified adult and pediatric SLE into two subgroups.SummarySeveral different endotypes of SLE patients with differing molecular profiles have been reported but a global consensus of clinically actionable groups has not been reached. Moreover, heterogeneity between datasets, reproducibility of predictions as well as the most effective classification approach have not been resolved. Nevertheless, gene expression-based precision medicine remains an attractive option to subset lupus patients.