Bayesian hierarchical modelling for inferring genetic interactions in yeast

Area of Science:

• Microbiology
• Computational Biology
• Genetics

Background:

• Quantitative fitness analysis (QFA) is a high-throughput method for assessing microbial population growth.
• QFA screens analyze gene mutations to infer genome-wide genetic interaction strengths across thousands of genotypes.

Purpose of the Study:

• To introduce advanced Bayesian hierarchical models for QFA that more accurately represent experimental design.
• To simultaneously model population dynamics and genetic interactions, improving upon current univariate summary methods.

Main Methods:

• Development of Bayesian hierarchical models for microbial population growth rates.
• Simultaneous modeling of population dynamics and genetic interactions within the QFA framework.
• Application of the new models to a published dataset for yeast telomere interactions.

Main Results:

• The Bayesian approach offers more efficient data utilization compared to existing methods.
• The models successfully identified new evidence for genes interacting with yeast telomeres.
• Simultaneous modeling avoids information loss associated with univariate fitness summaries.

Conclusions:

• Bayesian hierarchical models provide a more robust and efficient framework for quantitative fitness analysis.
• This methodology enhances the discovery of genetic interactions, particularly in complex biological systems like yeast telomeres.
• The improved approach better reflects QFA experimental designs, leading to more reliable inferences.