SSFold Software

Given the native structure of a protein, SSFold predicts the interaction order of secondary structure elements (SSEs) during folding. By considering each intermediate conformation as a collection of fully folded structures in which each of them contains a set of interacting SSEs, the conformation space is significantly reduced.

We show that SSFold is able to accurately predict the most energetically favorable folding pathway of large proteins with hundreds of residues at the mesoscopic level, including the pig muscle phosphoglycerate kinase with 416 residues. The model is detailed enough to distinguish between different folding pathways of structurally very similar proteins, including the streptococcal protein G and the peptostreptococcal protein L, and two variants NuG1 and NuG2 of protein G.

Using SSFold

The SSFold source code consists of a single file ssfold.c. It can be compiled under the Unix/Linux/Windows(Cygwin) environment with the command "gcc -O3 -o ssfold ssfold.c -lm".



ssfold -p 1GB1.pdb -s 1GB1SS -o 1GB1Fold

Command line parameters:

-p "name of file that contains the three-dimensional structure"
-s "name of file that contains the SSEs"
-o "name of output file"


Each row shows an intermediate conformation during folding. SSEs that are enclosed within a pair of parentheses are fully folded. The first number is the free energy of the conformation, and the second number is the percentage of native contacts of the conformation, where a native contact is defined to be a pair of amino acids that have their α-carbon atoms within 7 Å of each other.


   ( 1S )( 3S )( 5H )( 7S )( 9S )     43.94       0.50
   ( 1S )( 3S )( 5H )( 7S 9S )     21.43       0.61
   ( 1S )( 3S )( 5H 7S 9S )      0.29       0.70
   ( 1S 5H 7S 9S )( 3S )    -23.52       0.78
   ( 1S 3S 5H 7S 9S )    -66.40       1.00



Yang Q. and Sze S.-H. (2008) Predicting protein folding pathways at the mesoscopic level based on native interactions between secondary structure elements. BMC Bioinformatics, 9(320).