SiLiCO: A Simulator of Long Read Sequencing in PacBio and Oxford Nanopore

Abstract

Summary: Long read sequencing platforms, which include the widely used Pacific Biosciences (PacBio) platform and the emerging Oxford Nanopore platform, aim to produce sequence fragments in excess of 15-20 kilobases, and have proved advantageous in the identification of structural variants and easing genome assembly. However, long read sequencing remains relatively expensive and error prone, and failed sequencing runs represent a significant problem for genomics core facilities. To quantitatively assess the underlying mechanics of sequencing failure, it is essential to have highly reproducible and controllable reference data sets to which sequencing results can be compared. Here, we present SiLiCO, the first in silico simulation tool to generate standardized sequencing results from both of the leading long read sequencing platforms. Availability: SiLiCO is an open source package written in Python. It is freely available at

Technology and Species Independent Simulation of Sequencing Data and Genomic Variants

“…In this category, Grinder [15] has the unique feature to be able to accurately simulate shotgun and amplicon sequencing. The tools able to simulate long reads, instead, are usually specialized in either Pacific Bioscience SMRT sequencing or Oxford Nanopore sequencing with the notable exception of SiLiCO [16] that can simulate both of them. Simulators tailored on Pacific Bioscience SMRT sequencing are more common.…”

DeepSimulator: a deep simulator for Nanopore sequencing

“…The last module of DeepSimulator is the commonly used basecallers. Notice that during the entire simulating process, we do not explicitly introduce mismatches and indels (insertions and deletions), which is usually performed in the statistical simulators (Yang et al, 2017;Baker et al, 2016) directly at the read-level. Instead, we try to mimic the current signal produced by Nanopore sequencing as similar as possible, making the basecaller introduce mismatches and indels by itself.…”

“…As discussed in the previous papers (Yang et al, 2017;Baker et al, 2016), the read length of Nanopore sequencing is not very straightforward to model. Many factors, such as the experimental purpose and the experimenter's experience, would influence the read length distribution greatly.…”

“…1. Unlike the previous simulators (Yang et al, 2017;Baker et al, 2016) that only simulate the final reads from statistical models, our simulator attempts to mimic the entire pipeline of Nanopore sequencing. There are three main stages in Nanopore sequencing.…”

“…Despite the existence of more than twenty simulators for NGS technologies (Escalona et al, 2016), there are only three simulators created for the Nanopore sequencing, namely ReadSim (Lee et al, 2014), SiLiCO (Baker et al, 2016), and NanoSim (Yang et al, 2017). Although there are some differences between the three simulators (shown in Section S1), they share the same property of generating the simulated data utilizing the input nucleotide sequence and the explicit profiles 1 with a statistical model.…”