Pareto evolution of gene networks: an algorithm to optimize multiple fitness objectives

The computational evolution of gene networks functions like a forward genetic screen to generate, without preconceptions, all networks that can be assembled from a defined list of parts to implement a given function. Frequently networks are subject to multiple design criteria that cannot all be opti...

Full description

Saved in:
Bibliographic Details
Published in:Physical biology 2012-10, Vol.9 (5), p.056001-056001
Main Authors: Warmflash, Aryeh, Francois, Paul, Siggia, Eric D
Format: Article
Language:English
Subjects:
Algorithms
Biological Evolution
computational evolution
Feedback
Gene Regulatory Networks
Models, Genetic
pareto optimization
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The computational evolution of gene networks functions like a forward genetic screen to generate, without preconceptions, all networks that can be assembled from a defined list of parts to implement a given function. Frequently networks are subject to multiple design criteria that cannot all be optimized simultaneously. To explore how these tradeoffs interact with evolution, we implement Pareto optimization in the context of gene network evolution. In response to a temporal pulse of a signal, we evolve networks whose output turns on slowly after the pulse begins, and shuts down rapidly when the pulse terminates. The best performing networks under our conditions do not fall into categories such as feed forward and negative feedback that also encode the input-output relation we used for selection. Pareto evolution can more efficiently search the space of networks than optimization based on a single ad hoc combination of the design criteria.
ISSN:1478-3975
1478-3967
1478-3975
DOI:10.1088/1478-3975/9/5/056001