Fatal error: Too many LINCS warnings (1000) in a metadynamics simulation with CVs: COM distance and point of contacts

[PUJA BANERJEE]

Dear Plumed developers and users,

I am doing a metadynamics simulation of deaggregation of hexamer to tetramer+dimer using two CVs: distance between COM and pont of contacts.

This is my plumed script:

a: COM ATOMS=1-526,528-1053

b: COM ATOMS=1055-1580,1581-2106,2107-2632,2633-3158

d1:  DISTANCE ATOMS=a,b

d2: DISTANCES GROUPA=1-526,528-1053 GROUPB=1581-2106,2107-2632 LESS_THAN={RATIONAL R_0=0.2}

metad: METAD ARG=d1,d2.lessthan PACE=500 HEIGHT=2.0 SIGMA=0.01,0.01 FILE=HILLS BIASFACTOR=30.0 TEMP=300.0

PRINT STRIDE=100 ARG=d1,d2.lessthan,metad.bias FILE=COLVAR

I am getting an error:

Step 99812, time 199.624 (ps)  LINCS WARNING

relative constraint deviation after LINCS:

rms 0.080278, max 1.090188 (between atoms 253 and 254)

bonds that rotated more than 30 degrees:

 atom 1 atom 2  angle  previous, current, constraint length

    253    254  118.6    0.1141   0.2090      0.1000

    253    255   41.4    0.1234   0.0939      0.1000

...................

Fatal error:

Too many LINCS warnings (1000)

If you know what you are doing you can adjust the lincs warning threshold in your mdp file

or set the environment variable GMX_MAXCONSTRWARN to -1,

but normally it is better to fix the problem

For more information and tips for troubleshooting, please check the GROMACS

website at http://www.gromacs.org/Documentation/Errors

I tried with other width of gaussian : 0.005, 0.002, 0.001. everytime I am getting the same error after different times of simulation. An unbiased simulation with only GROMACS is running perfectly fine. This is my gromacs .mdp file:

title                    = NVT equilibration

;define                   = -DPOSRES

; RUN CONTROL PARAMETERS

integrator               = md

dt                       = 0.002   ; 2 femtosecond time step for integration

nsteps                   = 30000000  ; 1ns

; OUTPUT CONTROL OPTIONS

nstxout                  = 0    ; save coordinates every 2 ps

nstvout                  = 0    ; save velocities every 2 ps

nstenergy                = 2000    ; save energies every 2 ps

nstlog                   = 2000    ; update log file every 2 ps

nstxtcout                = 400

; NEIGHBORSEARCHING PARAMETERS

continuation             = yes

nstlist                  = 5

ns-type                  = Grid

pbc                      = xyz

rlist                    = 0.9

cutoff-scheme            =verlet

; OPTIONS FOR ELECTROSTATICS AND VDW

coulombtype              = pme

pmeorder                 = 4

rcoulomb                 = 1.2

fourierspacing           = 0.16

ewald_rtol               = 1e-5

optimize_fft             = yes

vdw-type                 = Cut-off

rvdw                     = 1.2

; Temperature coupling

Tcoupl                   = V-rescale

tc-grps                  = Protein      Non-Protein

tau_t                    = 0.1          0.1

ref_t                    = 300          300

; Freeze atoms

;freezegrps               = ZN

;freezedim                = Y Y Y

;energygrp-excl           = ZN ZN

; Pressure coupling

pcoupl                   = no

;Velocity generation

gen_vel                  = no

;gen_temp                = 300

;gen_seed                = -1

;Constrain all bonds

constraints              = h-bonds

constraint-algorithm     = Lincs

lincs_iter               = 2            ; accuracy of LINCS 

lincs_order              = 6            ; also related to accuracy 

If I am taking only COM distance as a CV, it is working perfectly fine.

 

can anybody help me with this, please? What should I change in LINCS algorithm details? or the problem lie within metadynamics gaussian hill height and width? 

With regards,

Puja