You are here: Home / Software / dcf


Design by Compatible Functions (dcf) is a peptide design principle for multifunctional peptides based on similarity to a functional reference class.

DCF is used to design multifunctional peptides by a two step process. Starting from a reference class of peptides DCF designs peptides the following way:

  1. It trains a machine learning predictor to recognize functional domains compatible with the reference peptides from where the user chooses a set of functional domains to be included in the design
  2. It joins the selected functional domains into a complete peptide while at the same time optimizing the peptide sequence to be indistinguishable frome the reference class.


Source code and installation

The exact source code used to generate the result presented in the paper can be downloaded here:

The code is currently only tested on Linux and you will need cmake to compile it. In order to compile under Ubuntu and Debian you can run the following to pull the code directly from the repository and build it:

sudo apt-get install cmake build-essential git 
cd ~/code
git clone  
cd dcf
mkdir bin && cd bin
cmake ..

For more information on installation and usage you can reference the instructions here:

« November 2017 »