Statistical analysis
We applied to sapply function to make descriptive statistics for clinical data. The Pearson correlations coefficients estimation between clinical data were evaluated by using cor function. Usingcor.test function examined the significance of correlation coefficients. TukeyHSD function was conducted to analyze the differences in average value of clinical data among groups.Fisher test is used to test the significance of frequency. Moreover, pie function was used to draw the percentage of all ABC mutations. All the analysis were preformed using R software. In addition, we used three web-available tools SIFT, Mutation Taster and FATHMM to predict protein damage. Predictions were defined as damaging, probable damaging, possible damaging, neutral when three in all prediction tools, two out of three, one out of three and none reached damaging, respectively.