Simulating sequences of the human genome with rare variants

Area of Science:

• Genetics
• Computational Biology
• Population Genetics

Background:

• Simulated samples are crucial for developing statistical methods to identify genetic variants linked to human diseases.
• Current simulation methods struggle to realistically model multi-locus selection and the fitness effects of new mutations.
• Natural selection significantly impacts the diversity and abundance of rare genetic variations in human populations.

Purpose of the Study:

• To develop a novel computer program for simulating large populations of gene sequences.
• To enable the simulation of multi-locus selection models with realistic fitness effect distributions.
• To generate more accurate datasets for studying genetic diseases associated with rare variants.

Main Methods:

• A forward-time simulation approach was employed to create the computer program.
• The program simulates multi-locus fitness schemes and single-locus selection models (random or locus-specific).
• It supports arbitrary quantitative trait or disease models, allowing for sample drawing and analysis.

Main Results:

• Simulated datasets using the new method showed significant differences in rare variant number and diversity compared to methods ignoring natural selection.
• Realistic demographic and natural selection models, derived from empirical data, were utilized.
• The developed program effectively simulates datasets with realistic rare genetic variant distributions.

Conclusions:

• The new simulation program provides a valuable tool for genetic research.
• It enables more accurate simulation of rare genetic variants, crucial for understanding genetic diseases.
• This advancement aids in the development of statistical methods for identifying disease-predisposing genetic variants.