Predicting SSRI-Resistance: Clinical Features and tagSNPs Prediction Models Based on Support Vector Machine

Area of Science:

• Neuroscience
• Psychiatry
• Genetics

Background:

• Recurrent major depressive disorder (MDD) often involves treatment resistance.
• Predicting individual antidepressant response is a significant clinical challenge.
• Combining clinical and biological data may improve treatment prediction.

Purpose of the Study:

• To develop predictive models for identifying patients resistant to Selective Serotonin Reuptake Inhibitors (SSRIs).
• To evaluate the utility of clinical variables and genetic markers in predicting SSRI resistance.

Main Methods:

• Utilized data from 857 patients with recurrent MDD, including 32 variables and 34 tagSNPs in the 5-HT signaling pathway.
• Employed machine learning, specifically Support Vector Machine (SVM), to build prediction models.
• Identified 12 significant clinical variables and 4 tagSNPs differentiating SSRI-resistant and non-resistant groups.

Main Results:

• Twelve clinical features (e.g., psychomotor retardation, psychotic symptoms, suicidality) were significantly associated with SSRI resistance (P<0.05).
• Ten SSRI-resistance predictive models were developed using SVM.
• Inclusion of tagSNP data improved the accuracy of the predictive models to some extent.

Conclusions:

• Data-driven machine learning identified 10 predictive models for SSRI resistance in MDD.
• These models may facilitate the prospective identification of patients likely to be resistant to SSRIs.
• Integrating genetic information with clinical data enhances prediction accuracy for antidepressant response.