Abstract
Mental health disorders are devastating illnesses, often misdiagnosed due to overlapping clinical symptoms. Among these conditions, bipolar disorder, schizophrenia, and schizoaffective disorder are particularly difficult to distinguish, as they share alternating positive and negative mood symptoms. Accurate and timely diagnosis of these diseases is crucial to ensure effective treatment and to tailor therapeutic management to each individual patient. In this context, it is essential to move beyond standard clinical assessment and employ innovative approaches to identify new biomarkers that can be reliably quantified. We previously identified a panel of RNA editing biomarkers capable of differentiating healthy controls from depressed patients and, among depressed patients, those with major depressive disorder and those with bipolar disorder. In this study, we integrated Adenosine-to-Inosine RNA editing blood biomarkers with clinical data through machine learning algorithms to establish specific signatures for bipolar disorder and schizophrenia spectrum disorders. This groundbreaking study paves the way for the application of RNA editing in other psychiatric disorders, such as schizophrenia and schizoaffective disorder. It represents a first proof-of-concept and provides compelling evidence for the establishment of an RNA editing signature for the diagnosis of these psychiatric conditions.