Abstract :
[en] Numerous complex networks composed of diverse interactions, either physical
interactions or functional associations, between macromolecules underlie most cellular functions.
The sets of physical and functional associations are respectively defined as biophysical and genetic
interactome networks. Mapping and characterizing biophysical and genetic interactome networks
is necessary, albeit not sufficient, to understand complex genotype to phenotype relationships.
However, as current individual interactome maps remain incomplete, their organization remains
mysterious and the relationships between distinct maps are unclear. Moreover, understanding the
nature of the interactions elucidated by each of these maps is essential to accurately interpret the
functional relevance of their integration. Saccharomyces cerevisiae is one of the few organisms for
which systematic genetic and biophysical maps have been generated at genome scale, making it
possible to compare them. For my PhD thesis, my colleagues and I focused on protein interactome
networks, and provided the first annotation of an expanded map of the yeast binary protein
interactome. We first assessed the coverage of the yeast binary interactome network by generating
an inventory of all protein-protein interactions reported in public repositories. Using
comprehensive experimental validations, we identified and selected the datasets with a majority of
binary direct interactions. Assembling a binary interactome network of only ~7,000 interactions
from the literature highlighted the imperative need to systematically expand the coverage of the
yeast binary interactome network. To that end, we expanded an available ORFeome collection, to
assemble a nearly complete collection, and used it to systematically test all possible protein pairs
to produce a new systematic binary map, YI-II. We revealed biological properties that govern the
organization of the cellular interactome by integrating the expanded yeast binary interactome map
with genetic network maps. Our results support recent observations, that the majority of
interactions in the interactome are likely of a different nature, with most being more transient,
potentially involved in context-specific regulatory processes. An understanding of the properties
that govern integration of genetic and biophysical maps, as provided by this study, would be key
to not only understand known genotype to phenotype relationships but identify novel ones.
