Cross-Comparison of Inflammatory Skin Disease Transcriptomics Identifies PTEN as a Pathogenic Disease Classifier in Cutaneous Lupus Erythematosus

The study used tissue transcriptomics to compare gene expression across nine inflammatory skin diseases, aiming to understand the molecular relationship between these diseases and identify disease-specific markers. The normalized datasets retained differential expression signals, enabling unsupervised disease clustering and the identification of disease-specific gene signatures. Using the NS-Forest algorithm, a minimal set of biomarkers was identified and validated for use as diagnostic disease classifiers. Notably, PTEN was identified as a specific marker for cutaneous lupus erythematosus (CLE) and was found to be strongly expressed by lesional keratinocytes in association with pathogenic type I interferons (IFNs). PTEN facilitated the type I IFN expression in keratinocytes by promoting nuclear translocation of phosphorylated IRF3. This suggests that cross-comparison of tissue transcriptomics can establish a molecular disease taxonomy and provide new pathogenic disease biomarkers.