Update announcement: the latest version of YeastNet (v3) is now available from here.

Networks: v. 2     Applications: Phenotype Prediction

  • YeastNet v. 2

    YeastNet v. 2 is a probabilistic functional gene network of yeast genes, constructed from ~1.8 million expermental observations from DNA microarrays, physical protein interactions, genetic interactions, literature, and comparative genomics methods. In total, YeastNet v.2 covers 102,803 linkages among 5,483 yeast proteins (95% of the validated proteome).

    YeastNet was constructed using a modified Bayesian integration of diverse data types, with each data type weighted according to how well it links genes that are known to share functions. Each interaction in YeastNet has an associated log-likelihood score (LLS) that measures the probability of an interaction representing a true functional linkage between two genes.

    YeastNet v. 2 reference:
    Lee, I., Li, Z. and Marcotte, E. M.. An improved, bias-reduced probabilistic functional gene network of baker's yeast, Saccharomyces cerevisiae (PLoS ONE 2007 2(10):e988). Download here.

    Download the network interactions:
    Benchmark set, common gene names (approx. 0.7MB).
    Benchmark set, systematic orf names (approx. 1.1MB).
    Full network, common gene names (approx. 2.9MB).
    Full network, systematic orf names (approx. 3.4MB).
    Evidence for each link, common gene names (approx. 7.8MB).
    Evidence for each link, systematic orf names (approx. 8.3MB).
    A web server to search the network interactively will be added shortly.

    Evidence in YeastNet v. 2 derives from the following datasets, listed in the order used in the evidence file (followed by the overall LLS score):

    Evidence code
    Co-citation of yeast genes
    Co-expression among yeast genes (500 microarray datasets)
    Gene neighbourhoods of bacterial and archaeal orthologs
    Yeast genetic interactions (multiple datasets)
    Literature curated yeast protein interactions
    Protein complexes from affinity purification/mass spectrometry (multiple datasets)
    Co-inheritance of bacterial orthologs of yeast genes
    Rosetta Stone protein-based functional linkages
    Protein interactions inferred from tertiary structures of complexes
    High-throughput yeast 2-hybrid assays (multiple datasets)

Network Applications

  • Network Based Phenotype Prediction

    We demonstrate that loss-of-function yeast phenotypes are predictable by guilt-by-association in functional gene networks. Testing 1,102 loss-of-function phenotypes from genome-wide assays of yeast reveals predictability of diverse phenotypes, spanning cellular morphology, growth, metabolism, and quantitative cell shape features. We apply the method to (1) extend a genome-wide screen by predicting, then verifying, genes whose disruption elongates yeast cells, and (2) predict human disease genes.

    Yeast network-based phenotype prediction reference:
    McGary, K. L., Lee, I., and Marcotte, E. M. Broad network-based predictability of Saccharomyces cerevisiae gene loss-of-function phenotypes (Genome Biology 2007 8(12):R258) Download here.

Questions/Comments: Email marcotte AT icmb dot utexas dot edu