Mathematical Model Development for Schizophrenia Risk Prediction Based on Assessment of Polymorphic Alleles Carriage in 13 Genetic Loci Affecting Pterin Metabolism

Abstract

Real-time PCR was used to analyze the carriage of alleles in 13 genetic loci that affect pterin metabolism in 116 patients with schizophrenia and 62 healthy volunteers. To analyze the accuracy of predicting the risk of schizophrenia, we used the binary logistic regression method with the assessment of the contribution of all studied loci. Results: A mathematical model was developed that makes it possible to predict the risk of schizophrenia manifestation in a carrier of the combination of genotypes MTHFD1 1958CC/MTRR 66GG with a probability of 90.6%, MTHFD1 1958CC/MTRR66AG with a probability of 81.9%. The use of this model is expedient in routine psychiatric practice among individuals at high risk of schizophrenia manifestation after replication in other samples and obtaining a larger volume of observations.