Upload
dangquynh
View
222
Download
4
Embed Size (px)
Citation preview
Bindingfreeenergy:Whatdidyoulearn?
Florentina TofoleanuNationalInstitutesofHealth
NationalHeart,Lung,andBloodInstitute
D3RWorkshop, LaJolla,March11th 2016
Freeenergycalculationprocedure
Weiteratedthebinding freeenergycalculationstep:- 3timesforCBClip systems- 3-12timesforOAH/OAMe systems
=0
+ ++ =0
Gilson,M.K.,Given,J.A,Bush,B.L.&McCammon,J.A.Biophys.J.72,10471069(1997)Boresch,S.etal.,J.Phys.Chem.B.95359551(2003)
TheDouble-DecouplingMethod(DDM)
Gbind = GrestonC + Gelec
C + GvdWC + Grestoff
C
GvdWM Gelec
M
IonicconcentrationNa3PO4 ionicconcentrationsusedexperimentally: 20mM @pH7.4forCBClip 10mM @pH11.5forG1-G5inOAH/OAMe G1-G5 50mM @pH11.5forG6inOAH/OAMe
Byusing theionicstrength,wetranslatedthemintoNaCl solutions: 50mM NaCl 25mM NaCl 165mMNaCl
TI+Docking*
*Submission withthelowestRMS
Under-estimatingenergies
Over-estimatingenergies
CBClip calculationsperformed byDr.Juyong Lee,NIH/NHLBI.
BAR+Docking
TI+MD
BAR+MD
G3:consistentlyover-estimating thebinding energy
What went right: G1
Predicted binding pose
Binding affinity (kcal/mol)
Experiment -5.84
Calculation (TI + DOCK) -7.61 (0.75)
What went right: G9
Predicted binding pose
Binding affinity (kcal/mol)
Experiment -7.40
Calculation (TI + DOCK) -7.85 (0.90)
What went wrong: G3
Predicted binding pose
Binding affinity (kcal/mol)
Experiment -4.02
Calculation (TI + DOCK) -9.78 (2.36)
What went wrong: G8
Predicted binding pose
Binding affinity (kcal/mol)
Experiment -6.18
Calculation (TI + DOCK) -2.71 (0.87)
G2
Docking result Gas phase simulation result
G4
Docking result Gas phase simulation result
G5
Docking result Gas phase simulation result
G6
Docking result Gas phase simulation result
CBClip hasanopenconformation.
G10
Docking result Gas phase simulation result
CBClip hasasemi-openconformation.
RMSerrorsofallCBClip submissionsSubmission # Method Name
1 CHARMM/CGENFF/BAR/TI/TIP3P/GALAXY-DOCK/DOCKING
2 SOMD/AM1BCC-GAFF-TIP3P/MBAR/A
3 CHARMM/CGENFF/TI/TIP3P/GALAXY-DOCK/DOCKING
4 CHARMM/CGENFF/BAR/TIP3P/GALAXY-DOCK/DOCKING
5 SOMD/AM1BCC-GAFF-TIP3P/MBAR/D
6 SOMD/AM1BCC-GAFF-TIP3P/MBAR/B
7 CHARMM/CGENFF/TI/TIP3P
8 CHARMM/CGENFF/BAR/TIP3P
9 MT-1 binding free energy
10 HSA/BEDAM/OPLS2005/AGBNP2
11 SOMD/AM1BCC-GAFF-TIP3P/MBAR/C
12 MT-2 binding free energy
CBClip calculationsperformed byDr.Juyong Lee,NIH/NHLBI.
CorrelationsforCBClip submissionsSubmission # Method Name
1 CHARMM/CGENFF/BAR/TI/TIP3P/GALAXY-DOCK/DOCKING
2 SOMD/AM1BCC-GAFF-TIP3P/MBAR/A
3 CHARMM/CGENFF/TI/TIP3P/GALAXY-DOCK/DOCKING
4 CHARMM/CGENFF/BAR/TIP3P/GALAXY-DOCK/DOCKING
5 SOMD/AM1BCC-GAFF-TIP3P/MBAR/D
6 SOMD/AM1BCC-GAFF-TIP3P/MBAR/B
7 CHARMM/CGENFF/TI/TIP3P
8 CHARMM/CGENFF/BAR/TIP3P
9 MT-1 binding free energy
10 HSA/BEDAM/OPLS2005/AGBNP2
11 SOMD/AM1BCC-GAFF-TIP3P/MBAR/C
12 MT-2 binding free energy
OAH&OAMe Bindingenergies
OAH/OAMe andguests
Host:charge-8Guests:charged&neutral
Ions:neutralized&ionicconcentration~Na3PO4
Parameters:ParamChem CGENFF
Explicitsolvent
Drycavity(nowatermoleculeswithin)
3-12calculationsforeachsystem
K.Vanommeslaeghe,E.Hatcher,C.Acharya,S.Kundu,S.Zhong,J.Shim,E.Darian,O.Guvench,P.Lopes, I.Vorobyov,A.D.MacKerell Jr.,, J.Comput.Chem. 2010, 31,671-690
Ligand
Protein
GalaxyDock:Protein-LigandDockingProgram
Goal:predictthebindingposeofagivenligandwhenitisboundtoagivenprotein
Sampling
Importantforbinding Cluestoimproving
affinity
Twomaincomponents
Scoring
W. -H. Shin, J. K. Kim, D. S. Kim, C. Seok, J. Comput. Chem. 2013, 34, 2647
GalaxyDock-HG:Host-GuestDocking
Selectthreebinding posesforeachsystem->simulatethelowest
NeutralG3&G5
2.9620*
Tetramethyl-ammonium-hydroxide
DistancefromQMcalculations;courtesyofFrankPickard
Assessingtheguestprotonationstateinthecomplex
+ +
00
+
+
#$%&$%'(
#$%&$%')
'*+,-./012+3
fromDDMcalculations
(fromQMpKa calculations)
fromDDMcalculations
G1,G2,G4,G6=negativelychargedG3,G5=neutral
BindingfreeenergysubmissionsforOAH
Binding freeenergieswereunder-estimated.
NoNMRdataforG4.
Bestsubmission:TI,RMSvalues=1.36/1.46kcal/mol
Differencebetweenexperimentalandeachcomputedvalue.
Binding freeenergieswereunder-estimated.HighesterrorsforG4.
ErrorsforBindingfreeenergyforOAH
Bestsubmission:TI,RMSvalues=1.36/1.46
BindingfreeenergysubmissionsforOAMe
NoITCdataforG4andG5.ExceptforG5,energieswereover-estimated.
Bestsubmission:BAR,RMS=0.87,1.94kcal/mol
Differencebetweenexperimentalandeachcomputedvalue.
Binding freeenergieswereover-estimated.HighesterrorsforG4.
ErrorsforBindingfreeenergyforOAMe
Bestsubmission:BAR,RMS=0.87,1.94kcal/mol
ExperimentalvsComputationalDataUnder-estimatingenergies
Over-estimatingenergies
Bindingenergiesweremostlyunder-estimated.
G6 G4
G5
G6
G5
G4hadconsistentlyhigherrorswithrespecttoexperiment.WesuspectthateithertheBrionneedsre-parametrization,orthesystemisneutral(unlikely).
Comparisonwithothers:OAH
Comparisonwithothers:OAMe
IncavityvsAtopthecavity
G=-10.541kcal/mol G=-6.4kcal/mol
OAH
WhydidwegettheworstresultsforG4?
Initialparametrization QMcalculation Lonepair addition*
TI:-7.84->-4.30(kcal/mol)BAR:-4.89->-3.84(kcal/mol)OAH
OAMe TI:-9.95->-0.08(kcal/mol)BAR:-10.54->1.55(kcal/mol)
Experimentalvalue:-9.38kcal/mol
Experimentalvalue:-2.38kcal/mol
*ParametersfromAlexMacKerells group
Analyzingonlythechargedspecies
Brisinsidethecavity->Gismorepositive
-4.30kcal/mol -3.97kcal/mol
-2.64kcal/mol-3.84kcal/mol
TI
BAR
OAH,G4+lonepair
Brisinsidethecavity->Gismorepositive
-0.08kcal/mol +2.10kcal/mol
+1.96kcal/mol+1.55kcal/mol
TI
BAR
OAMeG4+lonepair
Take-homemessages
CBClip:lowesterrorsforsmall/rigidguests
Parametrization dictated the open/closestateofCBClip
Ionicconcentrationisimportant
Treatmentofhalogens isimportant
Furtheranalysisofhydration
Acknowledgements Juyong LeeNIH/NHLBI FrankPickardNIH/NHLBI GerhardKnig NIH/NHLBI
JingHuang UniversityofMaryland Chaok Seok SeoulNationalUniversity
TimMiller NIH/NHLBI RichardVenableNIH/NHLBI
BernardBrooks