Yang Shen

[ Home | Research | Publications | CAPRI | CV | Links ]



Interconnected Cost Function Networks (iCFN): an efficient exact algorithm for multistate protein design with substate ensembles

Introduction

Interconnected Cost Function Networks (iCFN) is an efficient exact algorithm for computational protein design in which you can solve different problems such as:

  • Side chain packing

  • Protein design for a single objective such as stability

  • Binding affinity toward multiple targets

  • Multistate protein design with a single substate or multiple substates.

  • With its fast speed and low memory-usage, iCFN makes real-world large protein designs computationally tracatable or efficient, while guaranteeing the global optimum or sub-optimal solutions (sequences and substate structures). The code is not parallelized yet.

Download

  • Download the whole package of iCFN with binary code and sample data from here as well as the user manual from here (update: Nov. 8, 2018). Users can solve their own design problems with our code by supplying their own energy files (see details in the manual). Moreover, if you use OSPREY for energy calculations of each substate, we have provided scripts in the package to convert OSPREY energy files into our format.

  • Download sample data for designing specificity for T-cell receptor (TCR) in our Bioinformatics 2018 paper
    at positions 26, 28, 98, 100 data1, data2, data3 and data4.

Pre-requisites and Installation

  • iCFN was developed and tested in C with the compiler gcc 4.4.7 under CentOS 6.7 linux.

Usage

See user manual for details please.

Version

Binary code is version 3.0 (updated on Apr. 9 2017) with new scripts (updated on Oct. 21, 2018).

Reference

Mostafa Karimi and Yang Shen, Interconnected Cost Function Networks (iCFN): an efficient exact algorithm for multistate protein design, Bioinformatics 34(17), i811-i820.

Contact

Please contact: mostafa_karimi@tamu.edu or yshen@tamu.edu for any questions or comments.