Sim-it

Sim-it simulates structural variations and long-read sequencing reads (Nanopore and PacBio) to generate realistic datasets for benchmarking and comparing structural variant (SV) detection methods.

Key Features:

• Realistic Simulation Model: Generates simulations that reflect the distribution of structural variations within the human genome for controlled evaluation of SV analysis programs.
• Long-Read Data Simulation: Simulates long-read sequencing data for Nanopore and PacBio to assess the effects of sequencing technology and depth on SV detection.
• Benchmarking Capability: Provides realistic simulated datasets to benchmark and compare structural variant callers.
• Integration with GIAB Consortium Benchmarks: Aligns simulations with the Genome In A Bottle (GIAB) Consortium structural variation benchmark set.
• Development of combiSV Method: Outputs from Sim-it were used to develop combiSV, which combines results from five SV callers into an enhanced call set that improves recall and precision.

Scientific Applications:

• SV caller benchmarking: Evaluate performance metrics such as recall and precision of existing structural variant callers using controlled simulated datasets.
• Sequencing technology and depth assessment: Test the impact of Nanopore and PacBio sequencing technologies and varying sequencing depths on SV detection accuracy.
• Algorithm development and validation: Develop and refine algorithms and combined-call approaches for improved structural variation analysis, exemplified by combiSV.

Methodology:

Sim-it integrates realistic models of structural variations with simulation of long-read sequencing reads.