Protein-protein interaction probability landscapes for yeast replicative aging

We used probability landscapes to map the age-dependent protein-protein interaction profiles during Saccharomyces cerevisiae replicative aging. We used quasi-potentials, the negative logarithm of the probabilities, to visualize the downs and rises on the landscapes with respect to the young cells. When the interaction probability is gauged by protein abundance during aging, intra-essential protein interactions fell into a low-potential basin. When the interaction probability is estimated from gene expression, we observed that intra-essential protein interactions were located on a high-potential ridge. These opposite patterns coincide with the previously reported uncoupling of the transcriptome and proteome during yeast aging and were also observed in landscapes on different parameter spaces. A proteome-wide probability landscape revealed that the essential proteins favor the interactions with other essential proteins. The increases in essential hub protein abundances may partly explain the opposite trends observed in the landscapes.