PBSIM

PBSIM simulates PacBio single-molecule sequencing reads (continuous long reads, CLR, and circular consensus sequencing, CCS) to reproduce their length and error characteristics for de novo genome assembly and read-correction benchmarking.

Key Features:

• Read types: Simulates PacBio continuous long reads (CLR) and circular consensus sequencing (CCS) reads with characteristic length and error profiles.
• Read-length modeling: Models PacBio read lengths using a log-normal distribution identified in empirical data.
• Empirical parameterization: Parameterizes distributions and error characteristics based on analysis of 13 PacBio datasets.
• Generation methods: Implements both model-based and sampling-based methods to generate synthetic reads.
• Error profiles: Replicates the high error rate typical of CLRs and the lower error rate of CCS reads.
• Coverage simulation: Enables simulation at specified coverage depths to evaluate assembly performance (studies indicate CLR ≥15 and CCS ≥30 for extensive assemblies).

Scientific Applications:

• De novo genome assembly: Benchmarking and planning of de novo assembly strategies using simulated PacBio CLR and CCS reads.
• Hybrid error correction and assembly testing: Evaluating hybrid error-correction methods and assembly outcomes for PacBio sequencing data.
• Experimental design assessment: Optimizing sequencing coverage and assessing sequencing dynamics for PacBio experiments.

Methodology:

Analyzes 13 PacBio datasets to characterize read-length distributions (log-normal) and employs both model-based and sampling-based methods to generate simulated CLR and CCS reads.