CNVfam is a software package for jointly detecting copy number variations (CNV) in nuclear families genotyped using the Illumina platform. CNVfam has the ability to detect both inherited CNVs and de novo CNVs. Compared to algorithms designed for unrelated subjects, CNVfam substantially improves the power of detecting short CNVs carried by multiple family members. CNVfam is computationally very fast, taking less than one minute to analyze a trio and five minutes to analyze a nuclear family with five offsprings.
CNVfam detects CNVs in two steps. In the first step, CNVfam infers haplotype sharing pattern at each probe using high quality genotype data. In the second step, CNVfam jointly detects CNVs conditioning on the inferred haplotype sharing pattern by maximizing a penalized likelihood function. Once the haplotype sharing is inferred using SNP genotype data, CNVfam allows the detection of CNVs for subjects with poor intensity data (but with acceptable SNP data).
CNVfam is implemented using C++ and currently running in Linux platform. The package has been tested in Illumina 550K, 610K and 660K genotyping platforms. The package includes a readme file and a data example. Please contact Peng.Li2@nih.gov or Jianxin.Shi@nih.gov if you have questions.
Questions? Contact: Dr. Jianxin Shi