Error related to detecting number of atoms

Dear Lammps Users,

I am using pair style DPD and an NVE run. Though initially this detects the exact number of atoms in my simulation. Later during the NVE run the atoms are not detected

ERROR: Lost atoms: original 2100 current 2099 (thermo.cpp:385)

At step 3000 the job ends with this error.

Is this related to the distance between the atoms in my initial file. I performed a short minimization run before the NVE run and do not detect any error in this step.

2900 447.16953 0.086976995 0 670.52187 11.021052 85184
2950 445.95279 0.080544659 0 668.6912 10.990569 85184
3000 447.01355 0.086265324 0 670.28729 11.016921 85184
3050 452.64435 0.085870869 0 678.72908 11.155618 85184
3100 452.09878 0.087439163 0 677.91268 11.14316 85184
3150 451.13827 0.091938886 0 676.47711 11.118593 85184
3200 452.35056 0.095598817 0 678.29833 11.148255 85184
3250 452.64093 0.08069976 0 678.71877 11.155504 85184
3300 453.62441 0.11202014 0 680.22462 11.18104 85184
ERROR: Lost atoms: original 2100 current 2099 (thermo.cpp:385)

Thank you
John

See the error messages section of the manual for
many reasons why you can lose atoms. The answer
is almost always bad dynamics or boundary conditions.

Steve