AI algorithm combined with RNA editing-based blood biomarkers to discriminate bipolar from major depressive disorders in an external validation multicentric cohort

• Published in: Journal of affective disorders Vol. 356; pp. 385 - 393
• Main Authors: Salvetat, Nicolas, Checa-Robles, Francisco Jesus, Delacrétaz, Aurélie, Cayzac, Christopher, Dubuc, Benjamin, Vetter, Diana, Dainat, Jacques, Lang, Jean-Philippe, Gamma, Franziska, Weissmann, Dinah
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.07.2024; Elsevier
• Subjects: A-to-I RNA editing; Adult; Algorithms; Artificial Intelligence; Biochemistry, Molecular Biology; Biomarkers - blood; Bipolar disorder; Bipolar Disorder - blood; Bipolar Disorder - diagnosis; Blood biomarker; Cohort Studies; Computer Science; Depressive Disorder, Major - blood; Depressive Disorder, Major - diagnosis; Diagnosis, Differential; Female; Genomics; Human health and pathology; Humans; Life Sciences; Machine learning; Male; MDD; Middle Aged; Molecular biology; NGS; Psychiatric/Mental Health; Psychiatrics and mental health; Psychiatry diagnostic; Quantitative Methods; RNA Editing; Sensitivity and Specificity; Switzerland; Switzerland; Psychiatry diagnostic; Blood biomarker; Machine learning; Bipolar disorder; NGS; MDD; Artificial intelligence; A-to-I RNA editing; diagnostic; Psychiatry
• ISSN: 0165-0327; 1573-2517; 1573-2517
• DOI: 10.1016/j.jad.2024.04.022

Bipolar disorder (BD) is a leading cause of disability worldwide, as it can lead to cognitive and functional impairment and premature mortality. The first episode of BD is usually a depressive episode and is often misdiagnosed as major depressive disorder (MDD). Growing evidence indicates that peripheral immune activation and inflammation are involved in the pathophysiology of BD and MDD. Recently, by developing a panel of RNA editing-based blood biomarkers able to discriminate MDD from depressive BD, we have provided clinicians a new tool to reduce the misdiagnosis delay observed in patients suffering from BD. The present study aimed at validating the diagnostic value of this panel in an external independent multicentric Switzerland-based cohort of 143 patients suffering from moderate to major depression. The RNA-editing based blood biomarker (BMK) algorithm developped allowed to accurately discriminate MDD from depressive BD in an external cohort, with high accuracy, sensitivity and specificity values (82.5 %, 86.4 % and 80.8 %, respectively). These findings further confirm the important role of RNA editing in the physiopathology of mental disorders and emphasize the possible clinical usefulness of the biomarker panel for optimization treatment delay in patients suffering from BD. •The differential diagnosis of BD and MDD is challenging leading to a long delay in BD characterisation•RNA editing is an early regulatory system whose modifications are involved in psychiatric disorders like BD and MDD.•RNA editing related blood biomarkers are identified in patients with mood disorders•AI algorithm using RNA editing biomarkers will discriminate depressed BD from MDD and validated in two independant cohorts.