A Dynamic Pooling Approach to Extract Complete Allele Signal Information in Somatic Copy Number Alternations Detection

Accurately characterizing somatic copy number alterations (SCNAs) in cancers are of great importance in both deciphering tumorigenesis and progression and improving clinical diagnosis/treatment. Many computational methods in detecting SCNAs were proposed in recent years, and saas-CNV is among the best performers evaluated with empirical datasets. However, saas-CNV method inefficiently uses the allele dosage information in next-generation sequencing or microarray data. To this regard, we proposed and implemented a novel approach to extract the complete allele signal information for SCNA detection. Evaluated in an empirical dataset of hepatocellular carcinoma, we demonstrated the novel approach enhanced data signal-to-noise ratio, and resulted in improved detection of copy number alternations especially focal genome changes.