Baryshnikova Lab


Research areas


The yeast phenome

The budding yeast Saccharomyces cerevisiae has been the leading model organism for understanding the functional organization of biological systems. In particular, the yeast deletion collection has empowered hundreds of genome-wide phenotypic screens where mutant alleles of all ~5,000 nonessential genes in the yeast genome have been systematically tested for phenotypes of interest in various environmental conditions. Collectively, these data represent the diversity of the yeast physiological response to genetic perturbation and likely reflect the underlying functional relationships between genes, pathways and broader biological processes. To investigate these relationships systematically, we are compiling a global yeast phenome dataset composed of phenotypic data from all published systematic surveys of the yeast deletion collection (~200 papers, so far). These data include ~60 phenotypic traits (e.g., growth, morphology, specific metabolite and protein abundance and localization) and thousands of environmental conditions (e.g., chemical compounds, nutrient sources, expression of exogenous proteins, temperature, pH). In addition to published data, this global phenotypic compendium currently includes ~150 unpublished datasets, kindly provided by over 40 yeast laboratories. The final compendium of all phenotypes will be openly accessible online at, similar to the Saccharomyces Genome Database (SGD), but with quantitative or discrete values associated with each mutant.

Functional organization of a living cell

My former supervisor, Charlie Boone, showed this wonderful claymation movie during his acceptance speech for the Edward Novitski Prize at the 2014 Yeast Genetics Meeting in Seattle. My infinite thanks to Charlie and to his daughter Hannah Boone for taking the time to make this.

The advent of high-throughput genomic technologies, such as microarrays, yeast-2-hybrid (Y2H) and synthetic genetic array (SGA), enabled systematic analysis of large groups of genes and proteins from a multitude of different angles and under thousands of different experimental conditions. Over the last decade, these experiments produced an enormous collection of high-quality quantitative data that empowered unprecedented progress in the understanding of gene function. However, despite this incredible wealth of data, we are still lacking a clear understanding of how different datasets fit together to represent the functional organization of a living cell. The goal of our research is to use a simple unicellular eukaryote, such as budding yeast Saccharomyces cerevisiae, to understand the relationship between different types of genome-scale datasets and to construct a global unified model of a living cell, which would serve as a repository for our collective knowledge and a model for more complex cellular systems. Our work is largely based on biological network analysis and integration, and relies on the development of new data visualization tools and software.

Genome structure and evolution

The genetic linkage map of chromosome I in yeast Saccharomyces cerevisiae.
The genetic diversity of a population is largely promoted by meiotic recombination, a molecular mechanism that reshuffles parental genomes to generate new allelic combinations. While most genetic variants assort at random relative to their maternal and paternal configurations, a smaller portion of the genome tends to remain unchanged across generations due to physical and genetic linkage among loci. We are interested in understanding what is special about these loci and how recombination affects their evolutionary trajectories. We combine computational analysis of genome sequences with large scale genetic linkage data and other genome-wide maps of meiotic recombination to investigate the current structure of the yeast genome and understand factors that may have affected this structure during evolution.

Biological data visualization

The genetic landscape of yeast Saccharomyces cerevisiae. The network represents connections between genes with similar genetic interaction profiles (Costanzo, Baryshnikova, et al., Science, 2010).
Big Data, in biology and everywhere else, calls for big ideas for its analysis. More often than not, big ideas are generated by looking at the data, exploring their structure and uncovering natural ways to organize them. We are very interested in developing new strategies for data visualization, especially in the field of network biology and functional genomics, and aim at implementing successful approaches (i.e., those being currently used in research) as automatic tools.