This is an unfinished clean-room reimplementation of EVFold ( www.evfold.org )
Unfortunately Sanders/Marks distribute the DCA code [Martin Weigt's scientific contribution], but not the folding code that works on top of DCA output [their scientific contribution]. The only way to test if their method even works as described has been to try to reverse-engineer/reimplement it from what few SI files were made available. Recently they have taken down those files, but you can still find them on www.archive.org .
I am sorry about the state of this. If people really worked in an attitude of openness, it would never have been necessary in the firstplace.
You need Python to run the scripts. They are old and may only work on py2.7 . (sorry again)
You need the CNS package to do the computation. Licensing prevents me from providing it to you. [ https://www.mrc-lmb.cam.ac.uk/public/xtal/doc/cns/cns_1.3/installation/frame.html ]
There are scripts for CNS version 1.3 and 1.21 . I hope they both work.
Again, I would have liked to provide a docker image for this, but licensing prevents my redistributing CNS. :/
1) Environment
export PYTHONPATH=./utils/pymodules:${PYTHONPATH}
2) Weighted (.sim) input file.
./utils/SimpleHammWeight.py <input>.faa <output>.sim
Be careful, on huge alignments this may take quite a lot of time and memory.
Consider using hhblits to create your alignments and hhfilt to filter to get fewer sequences with more independence. That will speed things up a lot.
My favorite settings for hhblits are " -n 4 -nodiff -cov 75 -id 99 -nodiff -neffmax 20 --maxfilt 100000 "
My favorite settings for hhfilt are " hhfilter -cov 75 -id 95 -i <INTPUT> -o <OUTPUT>"
3)
consider seeing bin/prepare.bash and Makefile to see how to run the program.