[gmx-users] gmx covar error in gromacs2019.x

008 3325574168 at qq.com
Mon Jan 6 09:52:22 CET 2020


Hi gmx users!
I am using gmx2019.4 in centos 7.6 (gcc version 4.8.5), the MD was conducted successfully. However, when I proceeded the PCA analysis, there was a wrong in the output informations.


command:
echo 3 3|gmx covar -f md_res_center_mol_fit_protein.xtc -s ../md.tpr -o eigenvalues.xvg -v eigenvectors.trr -xpma covapic.xpm -n ../index.ndx


output:
Choose a group for the least squares fit
Group     0 (         System) has 124748 elements
Group     1 (        Protein) has  8851 elements
Group     2 (      Protein-H) has  4419 elements
Group     3 (        C-alpha) has   574 elements
Group     4 (       Backbone) has  1722 elements
Group     5 (      MainChain) has  2298 elements
Group     6 (   MainChain+Cb) has  2824 elements
Group     7 (    MainChain+H) has  2852 elements
Group     8 (      SideChain) has  5999 elements
Group     9 (    SideChain-H) has  2121 elements
Group    10 (    Prot-Masses) has  8851 elements
Group    11 (    non-Protein) has 115897 elements
Group    12 (          Other) has    23 elements
Group    13 (            MOL) has    23 elements
Group    14 (             NA) has     8 elements
Group    15 (          Water) has 115866 elements
Group    16 (            SOL) has 115866 elements
Group    17 (      non-Water) has  8882 elements
Group    18 (            Ion) has     8 elements
Group    19 (            MOL) has    23 elements
Group    20 (             NA) has     8 elements
Group    21 ( Water_and_ions) has 115874 elements
Group    22 (    Protein_lig) has  8874 elements
Group    23 (center_Val108_A) has    16 elements
Group    24 (        chain_A) has  4427 elements
Group    25 (        chain_B) has  4424 elements
Group    26 (C-alpha_&_chain_A) has   287 elements
Group    27 (C-alpha_&_chain_B) has   287 elements
Group    28 (            E53) has    15 elements
Group    29 (           K135) has    22 elements
Group    30 (           C169) has    11 elements
Group    31 (              S) has     1 elements
Group    32 (            C13) has     1 elements
Select a group: Selected 3: 'C-alpha'


Choose a group for the covariance analysis
Group     0 (         System) has 124748 elements
Group     1 (        Protein) has  8851 elements
Group     2 (      Protein-H) has  4419 elements
Group     3 (        C-alpha) has   574 elements
Group     4 (       Backbone) has  1722 elements
Group     5 (      MainChain) has  2298 elements
Group     6 (   MainChain+Cb) has  2824 elements
Group     7 (    MainChain+H) has  2852 elements
Group     8 (      SideChain) has  5999 elements
Group     9 (    SideChain-H) has  2121 elements
Group    10 (    Prot-Masses) has  8851 elements
Group    11 (    non-Protein) has 115897 elements
Group    12 (          Other) has    23 elements
Group    13 (            MOL) has    23 elements
Group    14 (             NA) has     8 elements
Group    15 (          Water) has 115866 elements
Group    16 (            SOL) has 115866 elements
Group    17 (      non-Water) has  8882 elements
Group    18 (            Ion) has     8 elements
Group    19 (            MOL) has    23 elements
Group    20 (             NA) has     8 elements
Group    21 ( Water_and_ions) has 115874 elements
Group    22 (    Protein_lig) has  8874 elements
Group    23 (center_Val108_A) has    16 elements
Group    24 (        chain_A) has  4427 elements
Group    25 (        chain_B) has  4424 elements
Group    26 (C-alpha_&_chain_A) has   287 elements
Group    27 (C-alpha_&_chain_B) has   287 elements
Group    28 (            E53) has    15 elements
Group    29 (           K135) has    22 elements
Group    30 (           C169) has    11 elements
Group    31 (              S) has     1 elements
Group    32 (            C13) has     1 elements
Select a group: Selected 3: 'C-alpha'
Calculating the average structure ...
Reading frame       0 time    0.000   
WARNING: number of atoms in tpx (574) and trajectory (8851) do not match
Last frame      10000 time 100000.000   


Back Off! I just backed up average.pdb to ./#average.pdb.3#
Constructing covariance matrix (1722x1722) ...
Last frame      10000 time 100000.000   
Read 10001 frames


Trace of the covariance matrix: 9.58344 (nm^2)


Back Off! I just backed up covapic.xpm to ./#covapic.xpm.3#
100%
Diagonalizing ...


-------------------------------------------------------
Program:     gmx covar, version 2019.4
Source file: src/gromacs/linearalgebra/eigensolver.cpp (line 137)


Fatal error:
Internal errror in LAPACK diagonalization.


For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors
-------------------------------------------------------




I had tested the same files in a virtual machine with centos 7.6 and gmx2019.4 installed, the output was normal and gave the output files I need. Then I used the same files on other server with gmx 4.5.3, and the md.tpr was replaced by md.gro, it was also normal. I had installed gmx2019.2, gmx2019.5 and it gave the same error.


the gmx was installed as follows:


cmake(version 3.16.0)

tar xvf cmake-*.tar.gz && cd cmake-*
./bootstrap
gmake && gmake install
yum remove cmake -y
ln -s /usr/local/bin/cmake /usr/bin/


fftw-3.3.8
tar -zxvf fftw* && cd fftw*
./configure --prefix=/usr/local/fftw3.3.8 --enable-sse2 --enable-avx  --enable-avx2 --enable-float --enable-shared
make install -j 16


gmx2019.4
    tar xfz gromacs-2019.4.tar.gz && cd gromacs-2019.4
    mkdir build && cd build
    export CMAKE_PREFIX_PATH=/usr/local/fftw3.3.8

    cmake .. -DCMAKE_INSTALL_PREFIX=/opt/software/gmx2019.4 -DREGRESSIONTEST_DOWNLOAD=ON -DGMX_GPU=ON -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-10.2
    make -j 16
    make check
    make install




Thanks in advance.
Regards


More information about the gromacs.org_gmx-users mailing list