Analysis methods for clinical databases which store genetic data

Following the advances in genomics and computer technologies, the role of bioinformatics in genetic studies dramatically increased. The need for new data analysis techniques to handle complex problems is one of the most challenging tasks that bioinformatics professionals have to deal with.Most frequent study design to compare allele frequency distances between patients and controls is called GWAS (Genome Wide Association Study). Since the first GWAS in 2005, increasing numbers of similar research were published. GWAS has its own strengths and weaknesses and still is a major design for most case ? control genetic research.One of the recent data analysis methods proposed for the analysis of genomic data is Multivariate Distance Matrix Regression (MDMR). MDMR is a statistically powerful tool, especially for the problems where the number of independent variables exceeds the total number of cases. A dependent variable matrix is calculated based on statistical distances between individuals and genetic information is added to the model as independent variables to explain the variation in the distance matrix. Until recently, its use in genetic research was limited with gene expression studies.This study utilizes traditional GWAS approach and MDMR method on a bipolar disorder data set. As the first step, we applied the GWAS technique to determine the statistically significant SNPs those may distinguish bipolar patients and control individuals. In the second step, we applied MDMR model to understand the phenotype X genotype interactions those may affect the disease course in bipolar disorder. As the last step, we compared the outcomes collected in previous steps and determined the overlapping SNPs.Despite several limitations, MDMR is a useful analysis tool, especially for investigating the effects of genetic factors on complex diseases. Flexibility of the method is an advantage as it can be used in both gene expression and SNP based genomic research. To our knowledge, this study is the first MDMR application on phenotype based distance matrices. Moreover, this is the first study to apply the method on bipolar disorder data.A total of 12 SNPs were common in both GWAS and MDMR runs. These SNPs are crucial in terms of both occurrence and course of bipolar disorder. Several SNPs and genes those found to be statistically significant in our analysis were previously reported to be associated with bipolar disorder.As a conclusion, MDMR is a useful tool and it should be tweaked in terms of its limitations so that it can be used more effectively in genomic research.