# Context:

Evolutionary trajectory that recovers a crippled phenotype towards a fully functional phenotype, by deleting (in a reproducible manner) other components of the network.

# Research questions:

1. Would be possible to predict this evolutionary trajectory, from existing knowledge?

   • Approach:
     • Using machine learning algorithms on existing publicly available datasets on budding yeast (fitness datasets, genetic /physical interactions, GO-terms) to decipher hidden patterns and relationships in the data that may explain this evolutionary trajectory, combined with biophysical modelling
   • State of the art: There is already some work on models to predict evolution based on biophysical knowledge, look here

2. What happens to the yeast interaction network in every step of the evolutionary trajectory ?

   • Approach :
     • Measure the essential genes and non essential genes beneficial for fitness for every step of the trajectory.
     • Translate the set of essential genes to the potential corresponding gene interaction network (computational approach)

3. What are the molecular mechanisms that lead to the adaptive changes for this trajectory?

   • Approach :
     • Computational approaches (network analysis ) feed with experimental data about essentiality/GO terms to decipher potential molecular mechanisms behind this trajectory , that can be used for further testing.

4. What can we learn from this trajectory that can we extend to others examples? (potential applications)