kalngyk / calibur

Source codes for the Calibur program reported in https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-11-25

This is the source codes for the Calibur program reported in https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-11-25

(For precompiled Windows binary of Calibur, please visit https://sourceforge.net/projects/calibur/)

How to compile:

If you are running a variant of UNIX (or MacOS/WSL/Cygwin/MSYS2), you should have the GNU C compiler (gcc) and Make (make) installed. At the directory where the file Makefile is, type make. This will build two programs: calibur and calibur-lite

• calibur is the main program.
• calibur-lite is a lighter version of calibur which reports only the best decoy.

If you are running MS Windows, you should have installed MS Visual Studio Community. Open the Developer Command Prompt at the directory where the file Makefile.win is, type nmake /F Makefile.win. This will build the program calibur.exe.

That's all!

How to run:

1. Prepare a file which list all the decoys' PDB files. That is, the file should read like:

decoysdir/1enhA0.pdb
decoysdir/1enhA1000.pdb
decoysdir/1enhA1001.pdb
decoysdir/1enhA1002.pdb
decoysdir/1enhA1003.pdb
...

Alternatively, you can use Rosetta's compose_score_silent.py to create a silent file out of your PDB files, and use this file as input to calibur.

2. Call calibur with this filename (mydecoys say) as the only parameter

   sh> calibur mydecoys (or sh> ./calibur mydecoys, or C:\> .\calibur mydecoys according to your system setup)

   calibur will figure out a reasonable threshold itself. You can also specify a threshold. For example, for threshold 1.0, run calibur as

   sh> calibur mydecoys 1.0

   You can also change the strategy that calibur uses to discover the threshold. To see how, run

   sh> calibur

   And the options will be shown.

calibur's Issues

calibur: command not found

After make, I type calibur but got calibur: command not found.
I also tried to type sh> calibur, and it just stuck here.
I use Ubuntu 20.04.4 LTS, if it helps.
btw, is calibur also apply to RNA structure cluster?

Segmentation fault (core dumped) on Linux and std::bad_alloc on Windows

On Windows:

C:\Users\11\Downloads\calibur.win32 (1)>calibur.exe -o pdblist.txt
Filtering on
Signature mode on
Using chains 'A', 'C', ' ' in PDB files
Read 60 decoy names
Estimating threshold range... complete
Estimated threshold range = [9999999.00000, 0.00000]
Most frequent distance = 9833330.00000
10.00000 percent distances below nan
Finding threshold to keep 35.93041% edges
Threshold = nan
Reading decoys with signature mode on and filter mode on
Read 60 decoys. Filtered 0 outlier decoys.
Decoys read in 0.01000 s
Initialized 60 decoys.
Use of SCUD's bound off
Using MATRIX mode
Auxiliary clustering... 98.3%
AuxiliaryClustering... completed in 0.02600 s
 completedcoys' neighbors... completed 98.3%
Finding and removing largest clusters...terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc

On Linux:

(base) :~/calibur-main$ ./calibur-lite /calibur-main/predicted_structures/pdblist.txt 
Filtering on
Signature mode on
Using chains 'A', 'C', ' ' in PDB files
Read 60 decoy names
Segmentation fault (core dumped)

Using calibur in Rosetta on Linux:

(base) :~$ $ROSETTA/calibur.linuxgccrelease -pdb_list /pdblist
core.init: Checking for fconfig files in pwd and ./rosetta/flags 
core.init: Rosetta version: rosetta.binary.linux.release-315 r315 2022.12+release.a4d7970 a4d79705213bc2acd6b51e370eddbb2738df6866 https://www.rosettacommons.org 2022-03-22T15:27:26.590656
core.init: command: /Rosetta/main/source/bin/calibur.linuxgccrelease -pdb_list /pdblist
basic.random.init_random_generator: 'RNG device' seed mode, using '/dev/urandom', seed=-678486154 seed_offset=0 real_seed=-678486154
basic.random.init_random_generator: RandomGenerator:init: Normal mode, seed=-678486154 RG_type=mt19937
core.init: Resolved executable path: /Rosetta/main/source/build/src/release/linux/5.4/64/x86/gcc/9/default/calibur.default.linuxgccrelease
core.init: Looking for database based on location of executable: /Rosetta/main/database/
Using C-alphas #1-end
Filtering on
Signature mode off
Using chains 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', ' ' in PDB files
Read 578 decoy names
Estimating threshold range...

AN INTERNAL ERROR HAS OCCURED. PLEASE SEE THE CONTENTS OF ROSETTA_CRASH.log FOR DETAILS.


Segmentation fault (core dumped)

I'm using calibur to clustering RNA structures.
pdblist.txt