Pysim-sv: a package for simulating structural variation data with GC-biases

Abstract

Background: Structural variations (SVs) are wide-spread in human genomes and may have important implications in disease-related and evolutionary studies. High-throughput sequencing (HTS) has become a major platform for SV detection and simulation serves as a powerful and cost-effective approach for benchmarking SV detection algorithms. Accurate performance assessment by simulation requires the simulator capable of generating simulation data with all important features of real data, such GC biases in HTS data and various complexities in tumor data. However, no available package has systematically addressed all issues in data simulation for SV benchmarking.

Results: Pysim-sv is a package for simulating HTS data to evaluate performance of SV detection algorithms. Pysim-sv can introduce a wide spectrum of germline and somatic genomic variations. The package contains functionalities to simulate tumor data with aneuploidy and heterogeneous subclones, which is very useful in assessing algorithm performance in tumor studies. Furthermore, Pysim-sv can introduce GC-bias, the most important and prevalent bias in HTS data, in the simulated HTS data.

Conclusions: Pysim-sv provides an unbiased toolkit for evaluating HTS-based SV detection algorithms.

Keywords: Breakpoints; Copy number variation; Next-generation sequencing; Translocation.

Fig. 1

The workflow of Pysim-sv. Component 1 simulates a personal genome by introducing genomic variations to a given reference genome. Component 2 generates tumor genomes by simulating aneuploidy and somatic variations. Subclones are iteratively generated. Component 3 generates HTS reads, mixes reads from different tumor/normal genomes and introduces GC-bias

Fig. 2

The GC-dependency in real data and simulated data. The GC-dependency in (a, b, c) three real sequencing data from the 1000 Genome Project and in (d, e, f) three simulated data generated by pysim-sv. The x-axis is the GC-proportion in 10 Kb bins and the y-axis is the number of mapped reads in the bins. Note that the lower bands in the left and middle panel of (a, b, c) correspond to bins in chromosome X and the two individuals here are two males. The functions in (d, e, f) are f ₁, f ₂ and f ₃ as presented in the GC-bias introduction section

Fig. 3

The sensitivity of the four SV detection algorithms with different parameters. Deletions (black), inversion (grey) and translocations (white) are compared, individually. a, b, and c are the simulated data with GC-bias, and d, e, f are the simulated data without GC-bias. The purities are 1 (a, d), 0.8 (b, e) and 0.5 (c, f)

Fig. 4

True CNVs in a simulated genome and detected by BIC-seq2. a Forty CNVs were introduced in the simulated genome and thirty-five copy number gains (red lines) and copy number losses (blue lines) were detected by BIC-Seq2. b True copy number gains (red lines) and copy number losses (blue lines) in the simulated genome