Figure 2. Identification of trait associated genes using robust regression adjusting for outlier effects.

(a) The top 10 genes with a large difference between ordinary and robust regression, which were visualized as fitted plot from the analysis of human BMI related RNA-seq data. (b) Dramatically different top 10 genes were visualized as fitted plot in the bovine RNA-seq data. In the (a,b), blue and red lines represent fitted line in the ordinary and robust regression, respectively, and the standard errors represent same color with fitted line. (c) Venn-diagram for comparing TAG list in each model with human RNA-seq data. (d) Venn-diagram to compare significantly detected gene list in the bovine RNA-seq data analysis. In the (c,d), raw P-value < 0.05 is used as cutoff for comparison among different methods.