[AMBER] mmgbsa tricky output

From: Amor San Juan <amorsanjuan.yahoo.com>
Date: Wed, 30 Jun 2010 02:19:53 -0700 (PDT)

I have protein (A), short peptide (B) and m7GTP (C). I calculated delta-H (enthalpy) using mmgbsa for different systems with results as follows:

1.) A + B =  -63.98
2.) A + C =  -49.34
3.)[A+B] + C = -48.72

Here is the tricky result when I tried calculating for [A+C] + B using the trajectory from (3) above, I got a value of -1686.27. Detailed values shown below:

[A+B] + C


                   COMPLEX               RECEPTOR                   LIGAND 
                     [A+B]+C                    A+B                        
   C    
#          ----------------------- ----------------------- -----------------------
#                  MEAN        STD         MEAN        STD         MEAN        STD
#          ======================= ======================= =======================
ELE            -7290.68      83.95     -6127.91      78.19      -175.61      11.69
VDW             -924.90      22.27      -890.49      21.70        -2.34       1.67
INT             4373.96      42.79      4256.68      42.01       117.68       7.44
GAS            -3841.62      89.97     -2761.72      85.19       -60.27      12.38
GBSUR             79.51       0.95        80.67       0.94         4.87       0.09
GB             -2967.91      71.26     -3318.90      69.23      -625.94      10.99
GBSOL          -2888.40      70.96     -3238.24      68.99      -621.07      10.93
GBELE         -10258.59      27.43     -9446.81      24.94      -801.55       5.06
GBTOT          -6730.02      46.04     -5999.96      44.23      -681.34       5.38

#                    DELTA       
#          -----------------------
#                  MEAN        STD
#          =======================
ELE             -987.16      61.60
VDW              -32.07       4.79
INT               -0.40       0.10
GAS            -1019.63      60.60
GBSUR             -6.03       0.19
GB               976.94      53.05
GBSOL            970.91      53.03
GBELE            -10.23      12.67
GBTOT            -48.72      11.11

[A+C]+B

#                  COMPLEX                RECEPTOR                  LIGAND
                     [A+C]+B                    A+C                           B      
#          ----------------------- ----------------------- -----------------------
#                  MEAN        STD         MEAN        STD         MEAN        STD
#          ======================= ======================= =======================
ELE            -7290.68      83.95     -5113.57      83.14      -247.41      15.01
VDW             -924.90      22.27      -845.52      21.31       -14.26       4.50
INT             4373.96      42.79      4051.41      41.25       255.06      10.75
GAS            -3841.62      89.97     -1907.69      89.32        -6.61      17.85
GBSUR             79.51       0.95        76.39       0.89        12.87       0.25
GB             -2967.91      71.26     -2705.33      73.00      -513.38      14.14
GBSOL          -2888.40      70.96     -2628.94      72.69      -500.52      14.03
GBELE         -10258.59      27.43     -7818.91      23.53      -760.79       4.10
GBTOT          -6730.02      46.04     -4536.63      42.73      -507.12      10.47

#                    DELTA       
#          -----------------------
#                  MEAN        STD
#          =======================
ELE            -1929.70      30.46
VDW              -65.12       4.04
INT               67.50       5.02
GAS            -1927.33      31.26
GBSUR             -9.75       0.33
GB               250.81      26.66
GBSOL            241.06      26.55
GBELE          -1678.89       9.39
GBTOT          -1686.27      10.80


Problem: As shown in red font, these are the values which is questionable or unrealistic.
My interpretation: ELE value is high which was not offset by small value of GB. Also, INT value is not close to zero  (single trajectory calculation). Note that structures of generated coordinates were all looking alright once checked/viewed in vmd.

Attempt to find "right' solution: Implemented igb=5 in mmgbsa, in contrast to igb=2 as used above. Would it make sense ? Try and see ... so below is the result:

[A+C]+B (using igb=5)

#                  COMPLEX                RECEPTOR                  LIGAND
                    [A+C]+B                   A+C                             B       
#          ----------------------- ----------------------- -----------------------
#                  MEAN        STD         MEAN        STD         MEAN        STD
#          ======================= ======================= =======================
ELE            -7290.68      83.95     -5113.57      83.14      -247.41      15.01
VDW             -924.90      22.27      -845.52      21.31       -14.26       4.50
INT             4373.96      42.79      4051.41      41.25       255.06      10.75
GAS            -3841.62      89.97     -1907.69      89.32        -6.61      17.85
GBSUR             79.51       0.95        76.39       0.89        12.87       0.25
GB             -1308.03      93.37     -1378.72      85.51      -421.59      13.52
GBSOL          -1228.52      93.02     -1302.33      85.18      -408.73      13.41
GBELE          -8598.71      72.13     -6492.30      61.17      -669.00       6.02
GBTOT          -5070.14      80.22     -3210.02      70.34      -415.33      11.10

#                    DELTA       
#          -----------------------
#                  MEAN        STD
#          =======================
ELE            -1929.70      30.46
VDW              -65.12       4.04
INT               67.50       5.02
GAS            -1927.33      31.26
GBSUR             -9.75       0.33
GB               492.29      33.91
GBSOL            482.54      33.82
GBELE          -1437.41      23.66
GBTOT          -1444.78      24.18

Alright, so as shown in blue font are values that changed upon using igb=5. So, the values isnt still "right".


I would greatly appreciate fresh ideas on how to find a "right" value for [A+C] +B system.

Thank you for all your time.

-amore-


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jun 30 2010 - 02:30:07 PDT
Custom Search