Quantitative Trait Specific Differential Expression (qtDE)

Natural selection based on phenotypic traits may lead to genetic changes. To understand the relationship between phenotypic traits and gene expression profiles it is important to study the concordance between the two. In this work we have developed a simple procedure to find the differentially expressed (DE) genes across various tissues between phenotypes through linear correlation as well as non linear mutual information and polynomial regression between quantitative-trait and the gene expression profiles. Here we are making the use of mice gene expression data to find the differentially expressed genes between the male and female phenotypes exploring the dependency between the gene expression profiles of four tissues (brain, muscle, liver, adipose) and quantitative trait (weight). To prove the effectiveness of the method we have tested our results with a popular DE tool (DEGseq). In the results we have shown that mutual information based trait-specific DE genes are biologically more significant compared to the polynomial regression and linear correlative counterparts.