A computational approach for the prediction of constitutive A-to-I editing sites in clusters from multiple RNA-seq datasets. STEP 1: Two-round unique mapping. STEP 2: Removal of sequencing errors and annotated gSNPs. STEP 3: Removal of unannotated gSNPs with customized PSS. STEP 4: Identification of constitutive A-to-I editing sites clustered in editing boxes. See Materials and Methods for details.