计算化学公社

标题: 求助amber QM-MM 最小化问题 [打印本页]

作者
Author: hhhnano 时间: 2020-7-25 20:01
标题: 求助amber QM-MM 最小化问题
本帖最后由 hhhnano 于 2020-7-26 18:06 编辑

我先生成prmtop 和inpcrd文件（有BOX坐标），运行MM-MD，一切正常，cpptraj处理周期性边界后，提取平衡时最后一帧（499帧）作为QM-MM输入PDB文件，重新生成prmtop 和inpcrd文件，发现inpcrd文件中的BOX坐标没有了，就将MM-MD运行前的inpcrd文件BOX坐标复制到新的inpcrd文件中，可以QM-MM最小化，但是在900、910步出现“... RESTARTED DUE TO LINMIN FAILURE ...”，计算终止，经过搜索百度，将“ntc = 2, ntf = 2,” 修改为“ntc = 1, ntf = 1, 将“ncyc = 600,”修改为“ncyc = 600,”，在940，970，990步时又出现这样的问题，但能计算结束，提示“ Maximum number of minimization cycles reached.“
请问这种结束结果对吗？如果不对，如何修改，非常感谢！

                  FINAL RESULTS

NSTEP    ENERGY       RMS          GMAX       NAME NUMBER
1000    -2.1398E+06    2.2966E-01    2.3479E+00    O       4518

BOND =    169.6754  ANGLE =    205.7311  DIHED    =    715.6707
VDWAALS =    -853.2027  EEL    =  -462140.5995  HBOND    =       0.0000
1-4 VDW =    254.7265  1-4 EEL =    3126.3753  RESTRAINT  =       0.0000
EXTERNESCF = -1681235.0114

--------------------------------------------------------------------------------
5.  TIMINGS
--------------------------------------------------------------------------------

|             Build the list          0.27 (42.71% of List )
|             Other                   0.37 (57.29% of List )
|          List time                0.64 ( 0.21% of Nonbo)
|                Short_ene time          85.69 (90.47% of Direc)
|                Other                   9.03 ( 9.53% of Direc)
|             Direct Ewald time       94.72 (31.38% of Ewald)
|             Adjust Ewald time       0.63 ( 0.21% of Ewald)
|                Fill Bspline coeffs       7.16 ( 3.48% of Recip)
|                Fill charge grid       17.53 ( 8.52% of Recip)
|                Scalar sum             80.79 (39.26% of Recip)
|                Grad sum                30.83 (14.98% of Recip)
|                FFT time                68.14 (33.11% of Recip)
|                Other                   1.32 ( 0.64% of Recip)
|             Recip Ewald time       205.77 (68.17% of Ewald)
|             Virial junk             0.40 ( 0.13% of Ewald)
|             Other                   0.29 ( 0.10% of Ewald)
|          Ewald time             301.84 (99.79% of Nonbo)
|       Nonbond force          302.49 ( 0.10% of Force)
|          Other                290498.79 (100.0% of QMMM )
|       QMMM                290513.69 (99.89% of Force)
|    Force time          290820.07 (100.0% of Runmd)
| Runmd Time          290820.07 (100.0% of Total)
| Total time          290832.85 (100.0% of ALL  )

| Number of list builds :       2

| Highest rstack allocated: 2181748
| Highest istack allocated:    35178
|          Job began  at 21:43:33.300  on 07/21/2020
|          Setup done at 21:43:33.937  on 07/21/2020
|          Run done at 06:30:46.150  on 07/25/2020
|    wallclock() was called 50064 times

|3D-RISM memory allocation summary
|Type       Maximum       Current
|Integer    0.00000 GB    0.00000 GB
|Real       0.00000 GB    0.00000 GB
|Logical    0.00000 GB    0.00000 GB
|Character    0.00000 GB    0.00000 GB
|---------------------------------------
|Total       0.00000 GB    0.00000 GB

最小化文件内容：
Initial min of our structure QMMM
&cntrl
  imin = 1, maxcyc = 1000, ncyc = 600,
  ntx = 1,
  irest = 0,
  ntb = 1,
  cut = 12.0, ntc = 1, ntf = 1,
  ntpr = 10,
  ntwx = 1,
  ibelly = 1,
  bellymask =':1 < @12.0',
  ifqnt = 1,
&end
&qmmm
qmmask=':1,12,15,122,294,295,8573',
  qmcharge = 0,
  spin = 2,
  qm_theory = 'EXTERN',
  qmshake= 0,
  qm_ewald=0, qm_pme=1,
  qmcut = 12.0,
  writepdb= 1
&end
&gau
executable = 'g16',
method = 'B3LYP',
basis = '6-31G*',
num_threads = 20,
mem = '200GB'
dipole = 1,
use_template = 0,
&end

作者
Author: chenjinfeng850 时间: 2020-7-26 20:59
检查一下输出的结构，如果结构没有什么问题，那么结果应该问题也不大

作者
Author: hhhnano 时间: 2020-7-27 10:31

chenjinfeng850 发表于 2020-7-26 20:59
检查一下输出的结构，如果结构没有什么问题，那么结果应该问题也不大

谢谢，昨天写信给AMBERLIST，和您说的一样，AMBER回复说计算正常。

Hi
RESTARTED DUE TO LINMIN FAILURE
This is not an error and needs no fixing. And looking at your FINAL
RESULTS, everything looks fine ok to me.

Best Regards
Elvis

欢迎光临计算化学公社 (http://bbs.keinsci.com/)