CNPI: Rapid Analyses of Human Copy Number Data

Abstract

Tools for genotyping copy number in whole-genome sequencing (WGS) data exist. Despite their availability, there are a limited number of tools that efficiently process this data at scale with rapid compute time, low memory usage, and easy to parse outputs. This is critical as WGS continues to be housed in the “cloud.” Copy Number Private Investigator (CNPI) is a computational toolkit (https://github.com/TNTurnerLab/CNPI) that performs several tasks including genotyping of regions of interest, digital karyotype determination, assessment of chromosomal sex, plotting of copy number in regions of the genome, and generation of an Individual-level Copy Number Score (ICNS). Inputs to CNPI include copy number estimates, genome-wide, from WGS data and a corresponding reference annotation file. With these inputs, CNPI can genotype any regions of the genome rapidly. For example, all genes in the genome can be genotyped in <2 s with 1 CPU and <1 GB memory. Benchmarking against gold-standard NA12878 data demonstrated good results with flexibility for users to optimize copy number thresholds based on desired sensitivity and specificity. Usage of CNPI is customizable with a variety of user-tailored functions and arguments. CNPI can be deployed to cloud based environments such as AWS lambda due to its intuitive design, near-zero cost per sample, and availability of compute resources. The application of ICNS has proven to be useful by comparing unaffected individuals to individuals with autism and 9p deletion syndrome, respectively. Overall, CNPI is advantageous for its ability to quickly provide reliable and precise copy number assessments of human genomic data.