Noise cancellation for robust copy number variation detection using next generation sequencing data

High-throughput next generation sequencing (NGS) technologies have created an opportunity for detecting copy number variations (CNVs) more accurately. However, efficient and precise detection of CNVs remains challenging due to high levels of noise and biases, data heterogeneity and the “big data” nature of NGS data. In this work, we introduce a novel preprocessing pipeline to improve the detection accuracy of CNVs in heterogeneous NGS data, such as cancer whole exome sequencing data. We employed several normalizations to reduce biases due to GC content, mappability and tumor contamination. We also developed a novel efficient and effective smoothing approach based on the Taut String method to reduce noise and increase the detection power of the CNV detection methods.