Major Depressive Disorder Diagnosis with miRNA-Based Quantum Machine Learning

Background and Objectives: Major Depressive Disorder (MDD) is a mental disorder characterized by persistent feelings of sadness, loss of interest, and agitation. MDD is the leading cause of disability worldwide, yet it is still largely underdiagnosed due to misinterpretation of behavioral symptoms. Despite growing evidence implicating microRNAs (miRNAs) in the regulation of genes linked to psychiatric disorders, few computational tools exist to reliably classify miRNAs as being associated with depression. This study presents a novel quantum machine learning approach to predict whether dysregulation of specific miRNAs is linked to depression. Methods: We used web scraping to create a dataset of sequence attributes, predicted gene target attributes, and predicted pathways attributes for 54 miRNAs frequently dysregulated in MDD patients and 54 randomly selected miRNAs. The dataset was limited to the most important attributes for classification using feature selection. Results: A quantum classifier was then built with 7 qubits, achieving an accuracy of 90%, and a Multilayer Perceptron model was built, achieving an accuracy of 87%. The role of cancer pathways was also implicated in MDD development. Discussion: This work represents one of the first applications of quantum machine learning to depression biomarker classification and provides proof of concept for its use in more complex mental health diagnostics. The quantum classifier not only handles entangled feature space and non-linear relationships to uncover patterns inaccessible to classical algorithms but also provides a real-world benchmark for using quantum computing in machine learning.