Background: The rapid development of next generation sequencing technology is changing the way of biological research in many aspects, which has become the most popular platform used for the genomic structural variation detection.The classical single-sample analysis methods make inference based on the sequencing error rate which is estimated from the base-calling/mapping quality scores.However, many factors during the sequencing data analysis pipeline may contribute to the sequencing errors but not be well considered by the quality scores, which cause genotype-call errors using these methods.