Fragment-based Computational Protein Design with Cost Function - - PowerPoint PPT Presentation

Fragment-based Computational Protein Design with Cost Function Network

David Simoncini, Sophie Barbe and Thomas Schiex

Math´ ematiques et Informatique appliqu´ ees de Toulouse, INRA

Simoncini & Barbe & Schiex Fragment-based CPD with CFN

Background

• PhD. Thesis, Computer Science

University of Nice-Sophia Antipolis, France Selective pressure in cellular evolutionary algorithms Postdoctoral fellowship Zhang IRU, RIKEN, Japan Fragment-based protein structure prediction Computational protein design

Simoncini & Barbe & Schiex Fragment-based CPD with CFN

Computational protein design and cost function networks

CPD as a cost function network Interactions can be decomposed as a sum of one-body and two-body energy terms.

ES =

n

• i=1

E(xr1

i ) + n−1

• i=1
• j>i

E(xr1

i , yr2 j )

Simoncini & Barbe & Schiex Fragment-based CPD with CFN

Fragment-based approach

Covering of a protein with fragments

Simoncini & Barbe & Schiex Fragment-based CPD with CFN