Difficulty running FoldX in linux

Question

[strayeroliver]

I am trying to complete a science fair project. I want to use FoldX and Boltz2 to find a stable mutation of PETase with cold affinity. I think I can do this by running FoldX to find stability (DDG) and Boltz2 for cold affinity on relevant mutations suggested by ChimeraX.

Specifically, by giving FoldX the 6eqe.pdb for PETase and running the RepairPDB and RunModel commands on all the different mutations of interest, I should get all their DDGs so that I can use the most stable ones for further study of cold affinities in Boltz2.

However, I am unable to get FoldX working properly. This is embarrassing because it's nearly the first step in the flow, and there are several papers written with Fold-X, so I know it must work. But, neither I, my mentor, nor my engineering father can figure it out. Any help, guidance, or mentorship would be much appreciated.

What doesn't work

I have 4 problems:

1. Version doesn't work. I can't seem to get the basic installation correct if the version doesn't even work, right?

2. Repaired" PDB is missing TER/END statements.

3. BuildModel: _FOLDXDEFINED error and wrong/no mutation implemented.

• Output message: Sorry yo specified an option_FOLDXDEFINED for the order of mutations that does not exist, should be _USERDEFINED or _FOLDXDEFINED

• No mutation applied (6EQE) or the wrong mutation is made (1UBQ: LEU at pos 8 mutated to ALA instead of VAL).

1. PositionScan: terminated: Specified residue not found for A8, empty PS_*.fxout.

What I did

1. Install FoldX 5.1

Download from the link sent after registration as a high school student - https://foldxsuite.crg.eu/system/files/foldx5_1Linux64.zip

Install it:

cd ~/Downloads
unzip foldx_1Linux64_0.zip -d /home/isef/proj/isef/tools/foldx/
cd /home/isef/proj/isef/tools/foldx/
chmod +x foldx_20251231
ls -l

total 199448
-rwxrwxr-x 1 isef isef  86720392 Apr 17 04:11 foldx_20251231
-rwxrwxr-x 1 isef isef 114894632 Apr 17 03:59 foldx4_20251231
drwxr-xr-x 2 isef isef      4096 Feb 23  2024 molecules
-rw-r--r-- 1 isef isef   2573870 Nov 15  2021 rotabase.txt
-rw-rw-r-- 1 isef isef     36791 Apr 17 03:59 yasaraPlugin.zip

1. Verify the binary

./foldx_20251231 -v 2>&1 | tee -a /tmp/foldx_version_log
   ********************************************
   ***                                      ***
   ***             FoldX 5.1 (c)            ***
   ***                                      ***
   ***     code by the FoldX Consortium     ***
   ***                                      ***
   ***     Jesper Borg, Frederic Rousseau   ***
   ***    Joost Schymkowitz, Luis Serrano   ***
   ***    Peter Vanhee, Erik Verschueren    ***
   ***     Lies Baeten, Javier Delgado      ***
   ***       and Francois Stricher          ***
   *** and any other of the 9! permutations ***
   ***   based on an original concept by    ***
   ***   Raphael Guerois and Luis Serrano   ***
   ********************************************

 *** THIS IS AN IMPORTANT MESSAGE: 
 >> External parametrized molecules were detected...
 >> Final energetic values will depend of the correct parametrization of this molecules. 


the required argument for option '--version' is missing

1. Run RepairPDB and check repaired file.

I started with 6eqe.pdb, my target enzyme, but that didn't work at all. By didn't "work at all", I mean that in addition to the 1UBQ errors to follow, it didn't perform any mutations at the desired position and reported 0s for total energy for all but the first run when given numberOfRuns > 1. So, I asked AI for the simplest possible mutation that FoldX can reliably do. It suggested mutating LEU at position 8 to VAL in 1ubq.pdb, so that's where I am focusing my effort. I just want to get 1 mutation to work correctly and then I can expand to my needs in 6eqe independently.

cd /tmp/foldx_temp
wget https://files.rcsb.org/download/1UBQ.pdb -O 1ubq.pdb
/home/isef/proj/isef/tools/foldx/foldx_20251231 --command=RepairPDB --pdb=1ubq.pdb --out-pdb=true 2>&1 | tee repair.log
grep -E "TER|END" 1ubq_Repair.pdb

The end of the repair says

End of Repair
Output PDB ./1ubq_Repair.pdb
Your file run OK
End time of FoldX: Mon Oct  6 19:32:47 2025
Total time spend: 35.80 seconds.
Cleaning RepairPDB...DONE

so I get my hopes up that it worked. Unfortunately, there are no TER or END in the repaired file.

1. Confirm LEU is at position 8 in the Amino Acid

$ awk '$6==8 {print}' 1ubq_Repair.pdb | head
ATOM     59  N   LEU A   8      30.132  40.069  18.642  1.00  9.84
ATOM     60  CA  LEU A   8      29.607  41.180  19.467  1.00 14.15
ATOM     61  C   LEU A   8      30.075  42.538  18.984  1.00 17.37
ATOM     62  O   LEU A   8      29.586  43.570  19.483  1.00 17.01
ATOM     63  CB  LEU A   8      29.919  40.890  20.938  1.00 16.63
ATOM     64  CG  LEU A   8      28.876  40.104  21.721  1.00 18.88
ATOM     65  CD1 LEU A   8      28.679  38.712  21.146  1.00 19.31
ATOM     66  CD2 LEU A   8      29.258  39.960  23.190  1.00 18.59

So, the 1ubq_Repair.pdb looks like it should at position 8.

1. Run BuildModel

5a. Create the individual_list.txt and BuildModel.cfg

cd /tmp/foldx_temp
echo "LA8V;" > individual_list.txt
cat > BuildModel.cfg << EOF
command=BuildModel
pdb=1ubq_Repair.pdb
mutant-file=individual_list.txt
numberOfRuns=1
out-pdb=true
EOF

5b. Run the BuildModel and save the log

/home/isef/proj/isef/tools/foldx/foldx -f BuildModel.cfg 2>&1 | tee /tmp/foldx_log
# Produced the following tail of output ...
Starting BuildModel
Reading LA8V;
Residue to Mutate LEUA8 has residue index 7
Sorry yo specified an option_FOLDXDEFINED for the order of mutations that does not exist, should be _USERDEFINED or _FOLDXDEFINED
Sorry yo specified an option_FOLDXDEFINED for the order of mutations that does not exist, should be _USERDEFINED or _FOLDXDEFINED
Your file run OK
End time of FoldX: Mon Oct  6 20:02:17 2025
Total time spend: 0.40 seconds.
validated file "1ubq_Repair_1.pdb" => successfully finished
Cleaning BuildModel...DONE

5c. Check to see if the mutation was done

awk '$6 == 8 {print $0}' /tmp/foldx_temp/1ubq_Repair_1.pdb
ATOM     59  N   ALA A   8      30.132  40.069  18.642  1.00  9.84
ATOM     60  CA  ALA A   8      29.607  41.180  19.467  1.00 14.15
ATOM     61  C   ALA A   8      30.075  42.538  18.984  1.00 17.37
ATOM     62  O   ALA A   8      29.586  43.570  19.483  1.00 17.01
ATOM     63  CB  ALA A   8      29.919  40.890  20.938  1.00 16.47

It's not LEU anymore, so that's good; however, it's ALA instead of VAL, which is not good.

5d. Check the output

$ cat Dif_1ubq_Repair.fxout 
FoldX 5.1 (2011)
by the FoldX Consortium 
Jesper Borg, Frederic Rousseau, Joost Schymkowitz,
Luis Serrano and Francois Stricher
------------------------------------------------------

PDB file analysed: batch
Output type: BuildModel
Pdb total energy    Backbone Hbond  Sidechain Hbond Van der Waals   Electrostatics  Solvation Polar Solvation Hydrophobic   Van der Waals clashes   entropy sidechain   entropy mainchain   sloop_entropy   mloop_entropy   cis_bond    torsional clash backbone clash  helix dipole    water bridge    disulfide   electrostatic kon   partial covalent bonds  energy Ionisation   Entropy Complex
1ubq_Repair_1.pdb   0.998679    0.00173548  0   0.867633    0   -0.917758   1.44122 -4.44089e-16    -0.361689   0.00608543  0   0   0   -0.0385523  -1.20861e-06    0   000 0   0   0

So, it seems to have produced total energy, but is this for that VAL mutation or the ALA mutation?

1. Run PositionScan.

Maybe we can figure get it with PositionScan by getting all the mutations at position 8 since specifying the mutation didn't work.

6a. Run it

/home/isef/proj/isef/tools/foldx/foldx --command=PositionScan --pdb=1ubq_Repair.pdb --positions=A8 --output-file=pos_scan 2>&1 | tee posscan.log
# Relevant tail of the output:
Specified residue not found.
no mutations performed.
Your file run OK
End time of FoldX: Mon Oct  6 20:13:45 2025
Total time spend: 0.60 seconds.
Cleaning PositionScan...DONE

6b. Check the output.

$ cat pos_scan 
terminated

6c. Are there any more files of note?

At the bottom of an ls -ltr there were only 3 files:

-rw-rw-r-- 1 isef isef     0 Oct  6 20:13 PS_pos_scan_scanning_output.txt
-rw-rw-r-- 1 isef isef    11 Oct  6 20:13 pos_scan
-rw-rw-r-- 1 isef isef  1804 Oct  6 20:13 posscan.log

Make logs available

repair.log for 1ubq
1ubq_Repair.pdb
foldx_log for 1ubq_Repair.pdb

[Mensur Dlakic]

The program is definitely working. It doesn't matter that it didn't produce TER and END lines at the end. Those are optional in PDB files.

For the rest, it is almost impossible to read this as the whole thing is improperly formatted. If you put three backtick signs (like this `, on the same key in the upper left corner as ~ or tilde) as a single line at the start of the output you want to format, and the same three backtick signs as a single line at the end, it will be formatted like this:

   ********************************************
   ***                                      ***
   ***             FoldX 5.1 (c)            ***
   ***                                      ***
   ***     code by the FoldX Consortium     ***
   ***                                      ***
   ***     Jesper Borg, Frederic Rousseau   ***
   ***    Joost Schymkowitz, Luis Serrano   ***
   ***    Peter Vanhee, Erik Verschueren    ***
   ***     Lies Baeten, Javier Delgado      ***
   ***       and Francois Stricher          ***
   *** and any other of the 9! permutations ***
   ***   based on an original concept by    ***
   ***   Raphael Guerois and Luis Serrano   ***
   ********************************************

the required argument for option '--version' is missing

Someone might be able to help from a properly formatted post.

By the way, a simple Google search will give you a clear clue how to fix this:

https://www.google.com/search?q=Sorry+you+specified+an+option_FOLDXDEFINED+for+the+order+of+mutations+that+does+not+exist%2C+should+be+_USERDEFINED+or+_FOLDXDEFINED&sca_esv=1e4238bc8dc008dd&rlz=1C1UEAD_enUS1016US1016&ei=Fv7laPewF6WQm9cPtaOg6QE&ved=0ahUKEwj3k8TC9pOQAxUlyOYEHbURKB0Q4dUDCBA&uact=5&oq=Sorry+you+specified+an+option_FOLDXDEFINED+for+the+order+of+mutations+that+does+not+exist%2C+should+be+_USERDEFINED+or+_FOLDXDEFINED

[strayeroliver]

Responding to your answer.

• I spent a lot of time formatting it just right, using the ``` methodology you mentioned. It looks very good on my end and in the vscode markdown renderer. I wonder what you see. How can I preview what you see if not the previewer here or in vscode?
• I made a lot of Google and AI searches before resorting to this forum. It's not simple at all for me. Rather, it's completely obtuse and even intractable without help.

Request:

• What I could really use, I think, is an example. Is anyone willing to share .cfg and .pdb files that work, as well as the output that they produce. I will feel much more confident once I get a single functioning run.

Answer 1

[Mensur Dlakic]

The post looks much better now, so you did something. The preview here works fine. It simply can't be that the preview for your original post looked fine to you.

Like I said, Google finds everything. I had no idea how to solve this problem, and it was in the Google output. I will allow that Google will not output exactly the same search results for all of us, depending on our browsing habits and location. Still, you should have been able to find the answer.

Removing Google from the equation, you could have found the answer by simply typing:

foldx -h

That's a generic way of asking for help in any Linux command (sometimes it is --help instead of -h). If you did that, in the resulting output there would be this section:

order of mutations:
  --order arg (=_FOLDXDEFINED)        Set the order in how we mutate several
                                      residues, _USERDEFINED (the user decides
                                      the order, _FOLDXDEFINED FoldX decides
                                      the order

So there you go, the order has to be specified as _USERDEFINED, which is what that warning (not an error) kept telling you. By the way, this answer was found by me after being suggested by Google and 10 minutes of my research time, so I wouldn't say the search results were obtuse at all.

The --order switch can be added to your BuildModel.cfg file by specifying an extra line (order=_USERDEFINED), in which case the command you used above should be fine.

cat > BuildModel.cfg << EOF
command=BuildModel
pdb=1ubq_Repair.pdb
mutant-file=individual_list.txt
order=_USERDEFINED
numberOfRuns=1
out-pdb=true
EOF

Also, it can be done by typing everything on the command line, assuming you still have the individual_list.txt file.

foldx -c BuildModel --pdb 1ubq_Repair.pdb --mutant-file individual_list.txt --order _USERDEFINED

In short, what you did originally worked but gave you an Alanine instead of Valine. When I ran the command as you did, the resulting file had the L -> A mutation, presumably because the _FOLDXDEFINED mutation order has a zero index and is specified differently. That part is for you to figure out, as the description must be somewhere in FoldX documentation or their website. By the way, you will probably want to customize the command above by adding the output PDB name.

Finally, the screen output of the command above, just to show that it has no warning and does the mutation as requested.

   ********************************************
   ***                                      ***
   ***             FoldX 5.1 (c)            ***
   ***                                      ***
   ***     code by the FoldX Consortium     ***
   ***                                      ***
   ***     Jesper Borg, Frederic Rousseau   ***
   ***    Joost Schymkowitz, Luis Serrano   ***
   ***    Peter Vanhee, Erik Verschueren    ***
   ***     Lies Baeten, Javier Delgado      ***
   ***       and Francois Stricher          ***
   *** and any other of the 9! permutations ***
   ***   based on an original concept by    ***
   ***   Raphael Guerois and Luis Serrano   ***
   ********************************************

1 models read: 1ubq_Repair.pdb
1 models read: 1ubq_Repair.pdb

BackHbond       =               -54.69
SideHbond       =               -21.17
Energy_VdW      =               -77.93
Electro         =               -2.42
Energy_SolvP    =               98.32
Energy_SolvH    =               -108.35
Energy_vdwclash =               2.63
energy_torsion  =               4.88
backbone_vdwclash=              31.88
Entropy_sidec   =               45.55
Entropy_mainc   =               102.83
water bonds     =               0.00
helix dipole    =               0.03
loop_entropy    =               0.00
cis_bond        =               0.00
disulfide       =               0.00
kn electrostatic=               0.00
partial covalent interactions = 0.00
Energy_Ionisation =             0.05
Entropy Complex =               0.00
-----------------------------------------------------------
Total          =                                  -10.28


BackHbond       =               -54.69
SideHbond       =               -21.17
Energy_VdW      =               -77.93
Electro         =               -2.42
Energy_SolvP    =               98.32
Energy_SolvH    =               -108.35
Energy_vdwclash =               2.63
energy_torsion  =               4.88
backbone_vdwclash=              31.88
Entropy_sidec   =               45.55
Entropy_mainc   =               102.83
water bonds     =               0.00
helix dipole    =               0.03
loop_entropy    =               0.00
cis_bond        =               0.00
disulfide       =               0.00
kn electrostatic=               0.00
partial covalent interactions = 0.00
Energy_Ionisation =             0.05
Entropy Complex =               0.00
-----------------------------------------------------------
Total          =                                  -10.28

Starting BuildModel
Reading LA8V;
Residue to Mutate LEUA8 has residue index 7
Mutating residue = LEUA8 into VAL
Your file run OK
End time of FoldX: Wed Oct  8 12:42:19 2025
Total time spend: 0.95 seconds.
validated file "1ubq_Repair_1.pdb" => successfully finished
Cleaning BuildModel...DONE