Re: [gmx-users] Charmm to gromacs conversion for lipids
Thank you Justin for the great help you have done, as you always been. I will download the latest updated CHARMM36 files. On Fri, Nov 7, 2014 at 6:10 PM, Justin Lemkul wrote: > > > On 11/7/14 12:30 AM, Venkat Reddy wrote: > >> Dear Justin, >> Please check the atom section of "toppar_all36_lipid_cholesterol.str" >> file >> that I pasted below. >> >> * Toppar stream file for cholesterol. Stream following reading of >> * top_all36_lipid.rtf and par_all36_lipid.rtf >> >> !reference >> !Suits, F., Pitman, M., MacKerell, A.D., Jr., Feller, S.E. "Molecular >> Level >> !Organization of Saturated and Polyunsaturated Fatty Acids in a >> !Phosphatidylcholine Bilayer Containing Cholesterol" Submitted >> !for publication >> >> > I see where the problem is. Note that this file says "Submitted for > publication," but it was actually published 10 years ago. This stream file > is outdated. The CHL1 model I included in the CHARMM36 distribution is the > current one, which uses CRL1 for C3 and a series of NBFIXes described in > the 2012 paper by Lim et al. (full reference in forcefield.doc). > > You can get the latest, updated CHARMM36 force field files from > http://mackerell.umaryland.edu/charmm_ff.shtml#charmm > > -Justin > > > read rtf card append >> * cholesterol residues >> * >> >> RESI CHL1 0.00 !cholesterol (name to avoid conflict with >> choline) >> ! atoms names correspond to the correct cholesterol nomenclature >> GROUP >> ATOM C3 CTL1 0.14 >> ATOM O3 OHL -0.66 >> ATOM H3' HOL 0.43 >> ATOM H3 HAL1 0.09 >> GROUP >> ATOM C4 CTL2-0.18 >> ATOM H4A HAL2 0.09 >> ATOM H4B HAL2 0.09 >> GROUP >> ATOM C5 CEL1 0.00 >> ATOM C6 CEL1-0.15 >> ATOM H6 HEL1 0.15 >> GROUP >> ATOM C7 CTL2-0.18 >> ATOM H7A HAL2 0.09 >> ATOM H7B HAL2 0.09 >> GROUP >> ATOM C8 CTL1-0.09 >> ATOM H8 HAL1 0.09 >> GROUP >> ATOM C14 CTL1-0.09 >> ATOM H14 HAL1 0.09 >> GROUP >> ATOM C15 CTL2-0.18 >> ATOM H15A HAL2 0.09 >> ATOM H15B HAL2 0.09 >> GROUP >> ATOM C16 CTL2-0.18 >> ATOM H16A HAL2 0.09 >> ATOM H16B HAL2 0.09 >> GROUP >> ATOM C17 CTL1-0.09 >> ATOM H17 HAL1 0.09 >> GROUP >> ATOM C13 CTL1 0.00 >> GROUP >> ATOM C18 CTL3-0.27 !methyl at c13 >> ATOM H18A HAL3 0.09 >> ATOM H18B HAL3 0.09 >> ATOM H18C HAL3 0.09 >> GROUP >> ATOM C12 CTL2-0.18 >> ATOM H12A HAL2 0.09 >> ATOM H12B HAL2 0.09 >> GROUP >> ATOM C11 CTL2-0.18 >> ATOM H11A HAL2 0.09 >> ATOM H11B HAL2 0.09 >> GROUP >> ATOM C9 CTL1-0.09 >> ATOM H9 HAL1 0.09 >> GROUP >> ATOM C10 CTL1 0.00 >> GROUP >> ATOM C19 CTL3-0.27 !methyl at c10 >> ATOM H19A HAL3 0.09 >> ATOM H19B HAL3 0.09 >> ATOM H19C HAL3 0.09 >> GROUP >> ATOM C1 CTL2-0.18 >> ATOM H1A HAL2 0.09 >> ATOM H1B HAL2 0.09 >> GROUP >> ATOM C2 CTL2-0.18 >> ATOM H2A HAL2 0.09 >> ATOM H2B HAL2 0.09 >> GROUP >> ATOM C20 CTL2-0.09 >> ATOM H20 HAL2 0.09 >> GROUP >> ATOM C21 CTL3-0.27 >> ATOM H21A HAL3 0.09 >> ATOM H21B HAL3 0.09 >> ATOM H21C HAL3 0.09 >> GROUP >> ATOM C22 CTL2-0.18 >> ATOM H22A HAL2 0.09 >> ATOM H22B HAL2 0.09 >> GROUP >> ATOM C23 CTL2-0.18 >> ATOM H23A HAL2 0.09 >> ATOM H23B HAL2 0.09 >> GROUP >> ATOM C24 CTL2-0.18 !beyond this nomenclature may not be correct >> ATOM H24A HAL2 0.09 >> ATOM H24B HAL2 0.09 >> GROUP >> ATOM C25 CTL1-0.09 !c25 >> ATOM H25 HAL1 0.09 >> GROUP >> ATOM C26 CTL3-0.27 !terminal methyl, c26 >> ATOM H26A HAL3 0.09 >> ATOM H26B HAL3 0.09 >> ATOM H26C HAL3 0.09 >> GROUP >> ATOM C27 CTL3-0.27 !terminal methyl, c27 >> ATOM H27A HAL3 0.09 >> ATOM H27B HAL3 0.09 >> ATOM H27C HAL3 0.09 >> >> On Thu, Nov 6, 2014 at 7:16 PM, Justin Lemkul wrote: >> >> >>> >>> On 11/6/14 8:38 AM, Venkat Reddy wrote: >>> >>> Dear Justin, Yes, the problem is with the blank line at the end of atomtypes.atp. Now its working fine. I have added the missing dihedral types from other sources. Thanks, I will fix that. >>> >>> I have compared the cholesterol CHARMM parameters that are converted to >>> gromacs format and original parameters from "toppar_all36_lipid_cholesterol.str" file available in CHARMM website. I found some variations in the atom types. For eg., The atom type of C3 atom from sterol ring in cholesterol is CTL1 in CHARMM whereas it is CG311 in charmm36-Nov2014.ff. Which residue are you looking at? There is no C3 atom with type CTL1 >>> in >>> that stream file. In any case, CG311 and CTL1 have identical parameters >>> (see ffnonbonded.itp) since CG311 is used for CGenFF model compounds, >>> which >>> in the case of lipids, were just ported over and given unique types. >>> Anything with a "G" as the second letter in the atom type is from CGenFF, >>> indicating "general" to di
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/7/14 12:30 AM, Venkat Reddy wrote: Dear Justin, Please check the atom section of "toppar_all36_lipid_cholesterol.str" file that I pasted below. * Toppar stream file for cholesterol. Stream following reading of * top_all36_lipid.rtf and par_all36_lipid.rtf !reference !Suits, F., Pitman, M., MacKerell, A.D., Jr., Feller, S.E. "Molecular Level !Organization of Saturated and Polyunsaturated Fatty Acids in a !Phosphatidylcholine Bilayer Containing Cholesterol" Submitted !for publication I see where the problem is. Note that this file says "Submitted for publication," but it was actually published 10 years ago. This stream file is outdated. The CHL1 model I included in the CHARMM36 distribution is the current one, which uses CRL1 for C3 and a series of NBFIXes described in the 2012 paper by Lim et al. (full reference in forcefield.doc). You can get the latest, updated CHARMM36 force field files from http://mackerell.umaryland.edu/charmm_ff.shtml#charmm -Justin read rtf card append * cholesterol residues * RESI CHL1 0.00 !cholesterol (name to avoid conflict with choline) ! atoms names correspond to the correct cholesterol nomenclature GROUP ATOM C3 CTL1 0.14 ATOM O3 OHL -0.66 ATOM H3' HOL 0.43 ATOM H3 HAL1 0.09 GROUP ATOM C4 CTL2-0.18 ATOM H4A HAL2 0.09 ATOM H4B HAL2 0.09 GROUP ATOM C5 CEL1 0.00 ATOM C6 CEL1-0.15 ATOM H6 HEL1 0.15 GROUP ATOM C7 CTL2-0.18 ATOM H7A HAL2 0.09 ATOM H7B HAL2 0.09 GROUP ATOM C8 CTL1-0.09 ATOM H8 HAL1 0.09 GROUP ATOM C14 CTL1-0.09 ATOM H14 HAL1 0.09 GROUP ATOM C15 CTL2-0.18 ATOM H15A HAL2 0.09 ATOM H15B HAL2 0.09 GROUP ATOM C16 CTL2-0.18 ATOM H16A HAL2 0.09 ATOM H16B HAL2 0.09 GROUP ATOM C17 CTL1-0.09 ATOM H17 HAL1 0.09 GROUP ATOM C13 CTL1 0.00 GROUP ATOM C18 CTL3-0.27 !methyl at c13 ATOM H18A HAL3 0.09 ATOM H18B HAL3 0.09 ATOM H18C HAL3 0.09 GROUP ATOM C12 CTL2-0.18 ATOM H12A HAL2 0.09 ATOM H12B HAL2 0.09 GROUP ATOM C11 CTL2-0.18 ATOM H11A HAL2 0.09 ATOM H11B HAL2 0.09 GROUP ATOM C9 CTL1-0.09 ATOM H9 HAL1 0.09 GROUP ATOM C10 CTL1 0.00 GROUP ATOM C19 CTL3-0.27 !methyl at c10 ATOM H19A HAL3 0.09 ATOM H19B HAL3 0.09 ATOM H19C HAL3 0.09 GROUP ATOM C1 CTL2-0.18 ATOM H1A HAL2 0.09 ATOM H1B HAL2 0.09 GROUP ATOM C2 CTL2-0.18 ATOM H2A HAL2 0.09 ATOM H2B HAL2 0.09 GROUP ATOM C20 CTL2-0.09 ATOM H20 HAL2 0.09 GROUP ATOM C21 CTL3-0.27 ATOM H21A HAL3 0.09 ATOM H21B HAL3 0.09 ATOM H21C HAL3 0.09 GROUP ATOM C22 CTL2-0.18 ATOM H22A HAL2 0.09 ATOM H22B HAL2 0.09 GROUP ATOM C23 CTL2-0.18 ATOM H23A HAL2 0.09 ATOM H23B HAL2 0.09 GROUP ATOM C24 CTL2-0.18 !beyond this nomenclature may not be correct ATOM H24A HAL2 0.09 ATOM H24B HAL2 0.09 GROUP ATOM C25 CTL1-0.09 !c25 ATOM H25 HAL1 0.09 GROUP ATOM C26 CTL3-0.27 !terminal methyl, c26 ATOM H26A HAL3 0.09 ATOM H26B HAL3 0.09 ATOM H26C HAL3 0.09 GROUP ATOM C27 CTL3-0.27 !terminal methyl, c27 ATOM H27A HAL3 0.09 ATOM H27B HAL3 0.09 ATOM H27C HAL3 0.09 On Thu, Nov 6, 2014 at 7:16 PM, Justin Lemkul wrote: On 11/6/14 8:38 AM, Venkat Reddy wrote: Dear Justin, Yes, the problem is with the blank line at the end of atomtypes.atp. Now its working fine. I have added the missing dihedral types from other sources. Thanks, I will fix that. I have compared the cholesterol CHARMM parameters that are converted to gromacs format and original parameters from "toppar_all36_lipid_cholesterol.str" file available in CHARMM website. I found some variations in the atom types. For eg., The atom type of C3 atom from sterol ring in cholesterol is CTL1 in CHARMM whereas it is CG311 in charmm36-Nov2014.ff. Which residue are you looking at? There is no C3 atom with type CTL1 in that stream file. In any case, CG311 and CTL1 have identical parameters (see ffnonbonded.itp) since CG311 is used for CGenFF model compounds, which in the case of lipids, were just ported over and given unique types. Anything with a "G" as the second letter in the atom type is from CGenFF, indicating "general" to differentiate them from the main CHARMM types. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For
Re: [gmx-users] Charmm to gromacs conversion for lipids
Dear Justin, Please check the atom section of "toppar_all36_lipid_cholesterol.str" file that I pasted below. * Toppar stream file for cholesterol. Stream following reading of * top_all36_lipid.rtf and par_all36_lipid.rtf !reference !Suits, F., Pitman, M., MacKerell, A.D., Jr., Feller, S.E. "Molecular Level !Organization of Saturated and Polyunsaturated Fatty Acids in a !Phosphatidylcholine Bilayer Containing Cholesterol" Submitted !for publication read rtf card append * cholesterol residues * RESI CHL1 0.00 !cholesterol (name to avoid conflict with choline) ! atoms names correspond to the correct cholesterol nomenclature GROUP ATOM C3 CTL1 0.14 ATOM O3 OHL -0.66 ATOM H3' HOL 0.43 ATOM H3 HAL1 0.09 GROUP ATOM C4 CTL2-0.18 ATOM H4A HAL2 0.09 ATOM H4B HAL2 0.09 GROUP ATOM C5 CEL1 0.00 ATOM C6 CEL1-0.15 ATOM H6 HEL1 0.15 GROUP ATOM C7 CTL2-0.18 ATOM H7A HAL2 0.09 ATOM H7B HAL2 0.09 GROUP ATOM C8 CTL1-0.09 ATOM H8 HAL1 0.09 GROUP ATOM C14 CTL1-0.09 ATOM H14 HAL1 0.09 GROUP ATOM C15 CTL2-0.18 ATOM H15A HAL2 0.09 ATOM H15B HAL2 0.09 GROUP ATOM C16 CTL2-0.18 ATOM H16A HAL2 0.09 ATOM H16B HAL2 0.09 GROUP ATOM C17 CTL1-0.09 ATOM H17 HAL1 0.09 GROUP ATOM C13 CTL1 0.00 GROUP ATOM C18 CTL3-0.27 !methyl at c13 ATOM H18A HAL3 0.09 ATOM H18B HAL3 0.09 ATOM H18C HAL3 0.09 GROUP ATOM C12 CTL2-0.18 ATOM H12A HAL2 0.09 ATOM H12B HAL2 0.09 GROUP ATOM C11 CTL2-0.18 ATOM H11A HAL2 0.09 ATOM H11B HAL2 0.09 GROUP ATOM C9 CTL1-0.09 ATOM H9 HAL1 0.09 GROUP ATOM C10 CTL1 0.00 GROUP ATOM C19 CTL3-0.27 !methyl at c10 ATOM H19A HAL3 0.09 ATOM H19B HAL3 0.09 ATOM H19C HAL3 0.09 GROUP ATOM C1 CTL2-0.18 ATOM H1A HAL2 0.09 ATOM H1B HAL2 0.09 GROUP ATOM C2 CTL2-0.18 ATOM H2A HAL2 0.09 ATOM H2B HAL2 0.09 GROUP ATOM C20 CTL2-0.09 ATOM H20 HAL2 0.09 GROUP ATOM C21 CTL3-0.27 ATOM H21A HAL3 0.09 ATOM H21B HAL3 0.09 ATOM H21C HAL3 0.09 GROUP ATOM C22 CTL2-0.18 ATOM H22A HAL2 0.09 ATOM H22B HAL2 0.09 GROUP ATOM C23 CTL2-0.18 ATOM H23A HAL2 0.09 ATOM H23B HAL2 0.09 GROUP ATOM C24 CTL2-0.18 !beyond this nomenclature may not be correct ATOM H24A HAL2 0.09 ATOM H24B HAL2 0.09 GROUP ATOM C25 CTL1-0.09 !c25 ATOM H25 HAL1 0.09 GROUP ATOM C26 CTL3-0.27 !terminal methyl, c26 ATOM H26A HAL3 0.09 ATOM H26B HAL3 0.09 ATOM H26C HAL3 0.09 GROUP ATOM C27 CTL3-0.27 !terminal methyl, c27 ATOM H27A HAL3 0.09 ATOM H27B HAL3 0.09 ATOM H27C HAL3 0.09 On Thu, Nov 6, 2014 at 7:16 PM, Justin Lemkul wrote: > > > On 11/6/14 8:38 AM, Venkat Reddy wrote: > >> Dear Justin, >> Yes, the problem is with the blank line at the end of atomtypes.atp. Now >> its working fine. I have added the missing dihedral types from other >> sources. >> >> > Thanks, I will fix that. > > I have compared the cholesterol CHARMM parameters that are converted to >> gromacs format and original parameters from >> "toppar_all36_lipid_cholesterol.str" file available in CHARMM website. I >> found some variations in the atom types. >> >> For eg., The atom type of C3 atom from sterol ring in cholesterol is CTL1 >> in CHARMM whereas it is CG311 in charmm36-Nov2014.ff. >> >> > Which residue are you looking at? There is no C3 atom with type CTL1 in > that stream file. In any case, CG311 and CTL1 have identical parameters > (see ffnonbonded.itp) since CG311 is used for CGenFF model compounds, which > in the case of lipids, were just ported over and given unique types. > Anything with a "G" as the second letter in the atom type is from CGenFF, > indicating "general" to differentiate them from the main CHARMM types. > > > -Justin > > -- > == > > Justin A. Lemkul, Ph.D. > Ruth L. Kirschstein NRSA Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 629 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > http://mackerell.umaryland.edu/~jalemkul > > == > -- > Gromacs Users mailing list > > * Please search the archive at http://www.gromacs.org/ > Support/Mailing_Lists/GMX-Users_List before posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/6/14 8:38 AM, Venkat Reddy wrote: Dear Justin, Yes, the problem is with the blank line at the end of atomtypes.atp. Now its working fine. I have added the missing dihedral types from other sources. Thanks, I will fix that. I have compared the cholesterol CHARMM parameters that are converted to gromacs format and original parameters from "toppar_all36_lipid_cholesterol.str" file available in CHARMM website. I found some variations in the atom types. For eg., The atom type of C3 atom from sterol ring in cholesterol is CTL1 in CHARMM whereas it is CG311 in charmm36-Nov2014.ff. Which residue are you looking at? There is no C3 atom with type CTL1 in that stream file. In any case, CG311 and CTL1 have identical parameters (see ffnonbonded.itp) since CG311 is used for CGenFF model compounds, which in the case of lipids, were just ported over and given unique types. Anything with a "G" as the second letter in the atom type is from CGenFF, indicating "general" to differentiate them from the main CHARMM types. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
Dear Justin, Yes, the problem is with the blank line at the end of atomtypes.atp. Now its working fine. I have added the missing dihedral types from other sources. I have compared the cholesterol CHARMM parameters that are converted to gromacs format and original parameters from "toppar_all36_lipid_cholesterol.str" file available in CHARMM website. I found some variations in the atom types. For eg., The atom type of C3 atom from sterol ring in cholesterol is CTL1 in CHARMM whereas it is CG311 in charmm36-Nov2014.ff. On Thu, Nov 6, 2014 at 6:21 PM, Justin Lemkul wrote: > > > On 11/6/14 12:45 AM, Venkat Reddy wrote: > >> Dear Justin, >> Thanks for the update. But unfortunately I couldn't find the parameters >> for >> cholesteryl oleate. So, I tried to convert again the parameters from >> CHARMM >> to GROMACS using these updated files: >> >> ./cgenff_charmm2gmx.py CHL1 CE-1.mol2 top_chol_ester.rtf >> /usr/local/gromacs/share/gromacs/top/charmm36-nov2014.ff >> >> Here I am getting an error like, >> >> Traceback (most recent call last): >>File "./cgenff_charmm2gmx.py", line 778, in >> atomtypes = read_gmx_atomtypes(atomtypes_filename) >>File "./cgenff_charmm2gmx.py", line 71, in read_gmx_atomtypes >> var = [entry[0],entry[1]] >> IndexError: list index out of range >> >> If I use the charmm36 files uploaded in March 2014, Its not throwing any >> error. Is there something wrong in the updated files? >> >> > Possibly. There is a blank line at the end of atomtypes.atp - does the > error go away if you delete this blank line? > > Also note that you don't need to re-convert your topology; what I was > suggesting was that the new force field files cover more parameters for > lipids, so if the needed dihedral is present (I can't actually say because > you only provided two of the atom types) then your topology should work > without any changes. > > > -Justin > > -- > == > > Justin A. Lemkul, Ph.D. > Ruth L. Kirschstein NRSA Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 629 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > http://mackerell.umaryland.edu/~jalemkul > > == > -- > Gromacs Users mailing list > > * Please search the archive at http://www.gromacs.org/ > Support/Mailing_Lists/GMX-Users_List before posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/6/14 12:45 AM, Venkat Reddy wrote: Dear Justin, Thanks for the update. But unfortunately I couldn't find the parameters for cholesteryl oleate. So, I tried to convert again the parameters from CHARMM to GROMACS using these updated files: ./cgenff_charmm2gmx.py CHL1 CE-1.mol2 top_chol_ester.rtf /usr/local/gromacs/share/gromacs/top/charmm36-nov2014.ff Here I am getting an error like, Traceback (most recent call last): File "./cgenff_charmm2gmx.py", line 778, in atomtypes = read_gmx_atomtypes(atomtypes_filename) File "./cgenff_charmm2gmx.py", line 71, in read_gmx_atomtypes var = [entry[0],entry[1]] IndexError: list index out of range If I use the charmm36 files uploaded in March 2014, Its not throwing any error. Is there something wrong in the updated files? Possibly. There is a blank line at the end of atomtypes.atp - does the error go away if you delete this blank line? Also note that you don't need to re-convert your topology; what I was suggesting was that the new force field files cover more parameters for lipids, so if the needed dihedral is present (I can't actually say because you only provided two of the atom types) then your topology should work without any changes. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
Dear Justin, Thanks for the update. But unfortunately I couldn't find the parameters for cholesteryl oleate. So, I tried to convert again the parameters from CHARMM to GROMACS using these updated files: ./cgenff_charmm2gmx.py CHL1 CE-1.mol2 top_chol_ester.rtf /usr/local/gromacs/share/gromacs/top/charmm36-nov2014.ff Here I am getting an error like, Traceback (most recent call last): File "./cgenff_charmm2gmx.py", line 778, in atomtypes = read_gmx_atomtypes(atomtypes_filename) File "./cgenff_charmm2gmx.py", line 71, in read_gmx_atomtypes var = [entry[0],entry[1]] IndexError: list index out of range If I use the charmm36 files uploaded in March 2014, Its not throwing any error. Is there something wrong in the updated files? On Thu, Nov 6, 2014 at 6:09 AM, Justin Lemkul wrote: > > > On 11/5/14 10:23 AM, Venkat Reddy wrote: > >> Dear Justin, >> >> While processing the lipid file using grompp, there was an error: >> >> ERROR 1 [file chl1.itp, line 938]: >>No default Proper Dih. types >> >> >> ERROR 2 [file chl1.itp, line 939]: >>No default Proper Dih. types >> >> >> ERROR 3 [file chl1.itp, line 940]: >>No default Proper Dih. types >> >> >> ERROR 4 [file chl1.itp, line 942]: >>No default Proper Dih. types >> >> >> ERROR 5 [file chl1.itp, line 944]: >>No default Proper Dih. types >> >> >> ERROR 6 [file chl1.itp, line 945]: >>No default Proper Dih. types >> >> >> ERROR 7 [file chl1.itp, line 946]: >>No default Proper Dih. types >> >> >> ERROR 8 [file chl1.itp, line 947]: >>No default Proper Dih. types >> >> >> ERROR 9 [file chl1.itp, line 948]: >>No default Proper Dih. types >> >> >> ERROR 10 [file chl1.itp, line 949]: >>No default Proper Dih. types >> >> >> ERROR 11 [file chl1.itp, line 956]: >>No default Proper Dih. types >> >> >> ERROR 12 [file chl1.itp, line 959]: >>No default Proper Dih. types >> >> >> I quickly went through the itp file and found atom types CEL1 and CTL1 as >> a >> part of dihedral are troubling. I have checked the parent rtf file for the >> dihedral definition of these atoms. The dihedral was defined as >> >> C3 C4 C5 C10 1.5367 111.70 54.04 115.09 1.5304 >> >> How to mention this information in ffbonded.itp? >> >> > Parameters should be specified by atom type, not atom name. Regardless, > you can always add parameters to [dihedraltypes] in ffbonded.itp after unit > conversion. Try the latest CHARMM36 force field files, which I just > uploaded today: > > http://mackerell.umaryland.edu/charmm_ff.shtml#gromacs > > I added a bunch of lipids and associated parameters that people had asked > for. Maybe what you need is already included there. > > -Justin > > -- > == > > Justin A. Lemkul, Ph.D. > Ruth L. Kirschstein NRSA Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 629 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > http://mackerell.umaryland.edu/~jalemkul > > == > -- > Gromacs Users mailing list > > * Please search the archive at http://www.gromacs.org/ > Support/Mailing_Lists/GMX-Users_List before posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/5/14 10:23 AM, Venkat Reddy wrote: Dear Justin, While processing the lipid file using grompp, there was an error: ERROR 1 [file chl1.itp, line 938]: No default Proper Dih. types ERROR 2 [file chl1.itp, line 939]: No default Proper Dih. types ERROR 3 [file chl1.itp, line 940]: No default Proper Dih. types ERROR 4 [file chl1.itp, line 942]: No default Proper Dih. types ERROR 5 [file chl1.itp, line 944]: No default Proper Dih. types ERROR 6 [file chl1.itp, line 945]: No default Proper Dih. types ERROR 7 [file chl1.itp, line 946]: No default Proper Dih. types ERROR 8 [file chl1.itp, line 947]: No default Proper Dih. types ERROR 9 [file chl1.itp, line 948]: No default Proper Dih. types ERROR 10 [file chl1.itp, line 949]: No default Proper Dih. types ERROR 11 [file chl1.itp, line 956]: No default Proper Dih. types ERROR 12 [file chl1.itp, line 959]: No default Proper Dih. types I quickly went through the itp file and found atom types CEL1 and CTL1 as a part of dihedral are troubling. I have checked the parent rtf file for the dihedral definition of these atoms. The dihedral was defined as C3 C4 C5 C10 1.5367 111.70 54.04 115.09 1.5304 How to mention this information in ffbonded.itp? Parameters should be specified by atom type, not atom name. Regardless, you can always add parameters to [dihedraltypes] in ffbonded.itp after unit conversion. Try the latest CHARMM36 force field files, which I just uploaded today: http://mackerell.umaryland.edu/charmm_ff.shtml#gromacs I added a bunch of lipids and associated parameters that people had asked for. Maybe what you need is already included there. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
Dear Justin, While processing the lipid file using grompp, there was an error: ERROR 1 [file chl1.itp, line 938]: No default Proper Dih. types ERROR 2 [file chl1.itp, line 939]: No default Proper Dih. types ERROR 3 [file chl1.itp, line 940]: No default Proper Dih. types ERROR 4 [file chl1.itp, line 942]: No default Proper Dih. types ERROR 5 [file chl1.itp, line 944]: No default Proper Dih. types ERROR 6 [file chl1.itp, line 945]: No default Proper Dih. types ERROR 7 [file chl1.itp, line 946]: No default Proper Dih. types ERROR 8 [file chl1.itp, line 947]: No default Proper Dih. types ERROR 9 [file chl1.itp, line 948]: No default Proper Dih. types ERROR 10 [file chl1.itp, line 949]: No default Proper Dih. types ERROR 11 [file chl1.itp, line 956]: No default Proper Dih. types ERROR 12 [file chl1.itp, line 959]: No default Proper Dih. types I quickly went through the itp file and found atom types CEL1 and CTL1 as a part of dihedral are troubling. I have checked the parent rtf file for the dihedral definition of these atoms. The dihedral was defined as C3 C4 C5 C10 1.5367 111.70 54.04 115.09 1.5304 How to mention this information in ffbonded.itp? Thank you for your valuable time. On Wed, Nov 5, 2014 at 6:16 PM, Venkat Reddy wrote: > Thank you Justin for the tip. I have given .rtf file as input and the > script has generated me the desired output .itp file. The command I used is > as shown below. > > ./cgenff_charmm2gmx.py CHL1 CE-1.mol2 top_chol_ester.rtf > /usr/local/gromacs/share/gromacs/top/charmm36.ff > > > Thank you very much Justin for the help. > > On Wed, Nov 5, 2014 at 6:05 PM, Justin Lemkul wrote: > >> >> >> On 11/5/14 7:33 AM, Venkat Reddy wrote: >> >>> Dear Justin, >>> My lipid is cholesteryl oleate (CO). Since there are CHARMM36 parameters >>> available for cholesterol and oleate chain in POPC, I have clubbed both >>> together to generate the topology for CO (rtf file). Now I want to >>> convert >>> this rtf to itp file. is there any shortcut to accomplish this task? >>> >>> >> In theory, our conversion script should handle that, because all the .str >> files are are .rtf and .prm entries, so if you don't have any new >> parameters, it can probably convert the .rtf itself. Never tried it, but >> you can probably make it work. Error messages are pretty descriptive in >> case of any issues. >> >> -Justin >> >> >> On Wed, Nov 5, 2014 at 5:57 PM, Justin Lemkul wrote: >>> >>> On 11/5/14 7:03 AM, Venkat Reddy wrote: Thank you Joao for the quick reply. My system has single lipid molecule > and > I have pre- calculated charge information for every atom in the lipid. > I > have edited charges on individual atoms apart from atom types. > By the way, Previous studies have reported a topology for same lipid > but > in > NAMD convention. Is there any other tool to convert NAMD to GROMACS > compatible format? > > > You have to be careful here. I would not use CGenFF for lipids, but for reasons different than what Joao was saying. As is cautioned in the CGenFF description, you shouldn't use it for molecules for which a highly tuned biomolecular force field already covers much of, if not all, of the chemical space. The CHARMM36 lipid force field is highly optimized and performs very well. CGenFF, by its nature, is generalized. With generalization and increased transferability comes a decrease in accuracy. Coupling charges calculated by a different method (is it compatible with what CHARMM normally requires, and are those charges tuned against the LJ parameters of the atom types to give reasonable intermolecular and water-water interactions?) with somewhat less accurate atom types and bonded parameters will probably lead to a poor result because you've got a hodge-podge of a topology. What is this lipid that you're trying to parametrize? Are its functional groups covered by the existing parameters in the CHARMM36 lipid force field? If they are, you should absolutely be using those parameters (charges, atom types, and bonded parameters). Even small inaccuracies or imbalances can have a huge impact, especially on lipids, which are very sensitive. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at h
Re: [gmx-users] Charmm to gromacs conversion for lipids
Thank you Justin for the tip. I have given .rtf file as input and the script has generated me the desired output .itp file. The command I used is as shown below. ./cgenff_charmm2gmx.py CHL1 CE-1.mol2 top_chol_ester.rtf /usr/local/gromacs/share/gromacs/top/charmm36.ff Thank you very much Justin for the help. On Wed, Nov 5, 2014 at 6:05 PM, Justin Lemkul wrote: > > > On 11/5/14 7:33 AM, Venkat Reddy wrote: > >> Dear Justin, >> My lipid is cholesteryl oleate (CO). Since there are CHARMM36 parameters >> available for cholesterol and oleate chain in POPC, I have clubbed both >> together to generate the topology for CO (rtf file). Now I want to convert >> this rtf to itp file. is there any shortcut to accomplish this task? >> >> > In theory, our conversion script should handle that, because all the .str > files are are .rtf and .prm entries, so if you don't have any new > parameters, it can probably convert the .rtf itself. Never tried it, but > you can probably make it work. Error messages are pretty descriptive in > case of any issues. > > -Justin > > > On Wed, Nov 5, 2014 at 5:57 PM, Justin Lemkul wrote: >> >> >>> >>> On 11/5/14 7:03 AM, Venkat Reddy wrote: >>> >>> Thank you Joao for the quick reply. My system has single lipid molecule and I have pre- calculated charge information for every atom in the lipid. I have edited charges on individual atoms apart from atom types. By the way, Previous studies have reported a topology for same lipid but in NAMD convention. Is there any other tool to convert NAMD to GROMACS compatible format? You have to be careful here. I would not use CGenFF for lipids, but >>> for >>> reasons different than what Joao was saying. As is cautioned in the >>> CGenFF >>> description, you shouldn't use it for molecules for which a highly tuned >>> biomolecular force field already covers much of, if not all, of the >>> chemical space. The CHARMM36 lipid force field is highly optimized and >>> performs very well. CGenFF, by its nature, is generalized. With >>> generalization and increased transferability comes a decrease in >>> accuracy. >>> Coupling charges calculated by a different method (is it compatible with >>> what CHARMM normally requires, and are those charges tuned against the LJ >>> parameters of the atom types to give reasonable intermolecular and >>> water-water interactions?) with somewhat less accurate atom types and >>> bonded parameters will probably lead to a poor result because you've got >>> a >>> hodge-podge of a topology. >>> >>> What is this lipid that you're trying to parametrize? Are its functional >>> groups covered by the existing parameters in the CHARMM36 lipid force >>> field? If they are, you should absolutely be using those parameters >>> (charges, atom types, and bonded parameters). Even small inaccuracies or >>> imbalances can have a huge impact, especially on lipids, which are very >>> sensitive. >>> >>> -Justin >>> >>> -- >>> == >>> >>> Justin A. Lemkul, Ph.D. >>> Ruth L. Kirschstein NRSA Postdoctoral Fellow >>> >>> Department of Pharmaceutical Sciences >>> School of Pharmacy >>> Health Sciences Facility II, Room 629 >>> University of Maryland, Baltimore >>> 20 Penn St. >>> Baltimore, MD 21201 >>> >>> jalem...@outerbanks.umaryland.edu | (410) 706-7441 >>> http://mackerell.umaryland.edu/~jalemkul >>> >>> == >>> -- >>> Gromacs Users mailing list >>> >>> * Please search the archive at http://www.gromacs.org/ >>> Support/Mailing_Lists/GMX-Users_List before posting! >>> >>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists >>> >>> * For (un)subscribe requests visit >>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or >>> send a mail to gmx-users-requ...@gromacs.org. >>> >>> >> >> >> > -- > == > > Justin A. Lemkul, Ph.D. > Ruth L. Kirschstein NRSA Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 629 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > http://mackerell.umaryland.edu/~jalemkul > > == > -- > Gromacs Users mailing list > > * Please search the archive at http://www.gromacs.org/ > Support/Mailing_Lists/GMX-Users_List before posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Li
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/5/14 7:33 AM, Venkat Reddy wrote: Dear Justin, My lipid is cholesteryl oleate (CO). Since there are CHARMM36 parameters available for cholesterol and oleate chain in POPC, I have clubbed both together to generate the topology for CO (rtf file). Now I want to convert this rtf to itp file. is there any shortcut to accomplish this task? In theory, our conversion script should handle that, because all the .str files are are .rtf and .prm entries, so if you don't have any new parameters, it can probably convert the .rtf itself. Never tried it, but you can probably make it work. Error messages are pretty descriptive in case of any issues. -Justin On Wed, Nov 5, 2014 at 5:57 PM, Justin Lemkul wrote: On 11/5/14 7:03 AM, Venkat Reddy wrote: Thank you Joao for the quick reply. My system has single lipid molecule and I have pre- calculated charge information for every atom in the lipid. I have edited charges on individual atoms apart from atom types. By the way, Previous studies have reported a topology for same lipid but in NAMD convention. Is there any other tool to convert NAMD to GROMACS compatible format? You have to be careful here. I would not use CGenFF for lipids, but for reasons different than what Joao was saying. As is cautioned in the CGenFF description, you shouldn't use it for molecules for which a highly tuned biomolecular force field already covers much of, if not all, of the chemical space. The CHARMM36 lipid force field is highly optimized and performs very well. CGenFF, by its nature, is generalized. With generalization and increased transferability comes a decrease in accuracy. Coupling charges calculated by a different method (is it compatible with what CHARMM normally requires, and are those charges tuned against the LJ parameters of the atom types to give reasonable intermolecular and water-water interactions?) with somewhat less accurate atom types and bonded parameters will probably lead to a poor result because you've got a hodge-podge of a topology. What is this lipid that you're trying to parametrize? Are its functional groups covered by the existing parameters in the CHARMM36 lipid force field? If they are, you should absolutely be using those parameters (charges, atom types, and bonded parameters). Even small inaccuracies or imbalances can have a huge impact, especially on lipids, which are very sensitive. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
Dear Justin, My lipid is cholesteryl oleate (CO). Since there are CHARMM36 parameters available for cholesterol and oleate chain in POPC, I have clubbed both together to generate the topology for CO (rtf file). Now I want to convert this rtf to itp file. is there any shortcut to accomplish this task? On Wed, Nov 5, 2014 at 5:57 PM, Justin Lemkul wrote: > > > On 11/5/14 7:03 AM, Venkat Reddy wrote: > >> Thank you Joao for the quick reply. My system has single lipid molecule >> and >> I have pre- calculated charge information for every atom in the lipid. I >> have edited charges on individual atoms apart from atom types. >> By the way, Previous studies have reported a topology for same lipid but >> in >> NAMD convention. Is there any other tool to convert NAMD to GROMACS >> compatible format? >> >> > You have to be careful here. I would not use CGenFF for lipids, but for > reasons different than what Joao was saying. As is cautioned in the CGenFF > description, you shouldn't use it for molecules for which a highly tuned > biomolecular force field already covers much of, if not all, of the > chemical space. The CHARMM36 lipid force field is highly optimized and > performs very well. CGenFF, by its nature, is generalized. With > generalization and increased transferability comes a decrease in accuracy. > Coupling charges calculated by a different method (is it compatible with > what CHARMM normally requires, and are those charges tuned against the LJ > parameters of the atom types to give reasonable intermolecular and > water-water interactions?) with somewhat less accurate atom types and > bonded parameters will probably lead to a poor result because you've got a > hodge-podge of a topology. > > What is this lipid that you're trying to parametrize? Are its functional > groups covered by the existing parameters in the CHARMM36 lipid force > field? If they are, you should absolutely be using those parameters > (charges, atom types, and bonded parameters). Even small inaccuracies or > imbalances can have a huge impact, especially on lipids, which are very > sensitive. > > -Justin > > -- > == > > Justin A. Lemkul, Ph.D. > Ruth L. Kirschstein NRSA Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 629 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > http://mackerell.umaryland.edu/~jalemkul > > == > -- > Gromacs Users mailing list > > * Please search the archive at http://www.gromacs.org/ > Support/Mailing_Lists/GMX-Users_List before posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/5/14 7:03 AM, Venkat Reddy wrote: Thank you Joao for the quick reply. My system has single lipid molecule and I have pre- calculated charge information for every atom in the lipid. I have edited charges on individual atoms apart from atom types. By the way, Previous studies have reported a topology for same lipid but in NAMD convention. Is there any other tool to convert NAMD to GROMACS compatible format? You have to be careful here. I would not use CGenFF for lipids, but for reasons different than what Joao was saying. As is cautioned in the CGenFF description, you shouldn't use it for molecules for which a highly tuned biomolecular force field already covers much of, if not all, of the chemical space. The CHARMM36 lipid force field is highly optimized and performs very well. CGenFF, by its nature, is generalized. With generalization and increased transferability comes a decrease in accuracy. Coupling charges calculated by a different method (is it compatible with what CHARMM normally requires, and are those charges tuned against the LJ parameters of the atom types to give reasonable intermolecular and water-water interactions?) with somewhat less accurate atom types and bonded parameters will probably lead to a poor result because you've got a hodge-podge of a topology. What is this lipid that you're trying to parametrize? Are its functional groups covered by the existing parameters in the CHARMM36 lipid force field? If they are, you should absolutely be using those parameters (charges, atom types, and bonded parameters). Even small inaccuracies or imbalances can have a huge impact, especially on lipids, which are very sensitive. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
On 11/5/14 5:28 AM, João Martins wrote: The rtp file is used by pdb2gmx for creating GROMACS topology and coordinate files. Changing it in the lipids.rtp will not do much other than allow you to generate from a pdb the .gro and .top for that molecule. Changing atom types is tricky and should be avoided in order to make sure you're using the cgenff-generated parameters, those atomtypes are connected to the molecule you submitted and the charmm36.ff used in the conversion. I'd advise against using cgenff-derived parameters for serious calculations, the charge penalties are usually pretty high and that means you won't have a realistic description of your system. However, if you In fairness, though, the topology produced by CGenFF tells users what magnitude of penalty is considered acceptable and what magnitude requires the user to do a bit of work on their own to improve the quality of the topology. Anyone that blindly uses a topology with high penalties...well, the quality of what comes out depends on the quality of what goes in :) CGenFF parameters can be used for "serious calculations," as we do all the time, provided the user understands what the limitations are and what additional work might be expected. It's impossible for us to cover all chemical space, so some of the analogies assigned by the CGenFF program are inherently poor, but are the "best guess" for that particular molecule, within the caveat that we can't do everything yet (if ever). -Justin really want to run a simulation using those parameters, that's perfectly possible. That command generates a usable topology and coordinate pair, as well as a parameter file that's loaded in the topology. For all intents and purposes, using those for a single molecule simulation will work and should be relatively straightforward. If you want to simulate a multi-molecule system, you can either manually create a coordinate file modifying the generated one or, alternatively, generate it using any molecule editing software, exporting a pdb and using pdb2gmx to generate a gromacs-compatible coordinate and topology file. *Joao Martins* joaomartins...@gmail.com On Wed, Nov 5, 2014 at 10:31 AM, Venkat Reddy wrote: Dear all, I have a new lipid molecule and I want to simulate it using charmm36 forcefield. I have generated .str file using *paramchem* server and then using the following command, I got the itp file. ./cgenff_charmm2gmx.py COL COL.mol2 COL.str charmm36.ff Is this a reasonable approach for generating gromacs topology for a lipid molecule? If I change the atom types in the generated itp file to lipid atom types described in lipids.rtp, is it appropriate? Thank you for your time -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
Thank you Joao for the quick reply. My system has single lipid molecule and I have pre- calculated charge information for every atom in the lipid. I have edited charges on individual atoms apart from atom types. By the way, Previous studies have reported a topology for same lipid but in NAMD convention. Is there any other tool to convert NAMD to GROMACS compatible format? Thanks alot On Wed, Nov 5, 2014 at 3:58 PM, João Martins wrote: > The rtp file is used by pdb2gmx for creating GROMACS topology and > coordinate files. Changing it in the lipids.rtp will not do much other than > allow you to generate from a pdb the .gro and .top for that molecule. > Changing atom types is tricky and should be avoided in order to make sure > you're using the cgenff-generated parameters, those atomtypes are connected > to the molecule you submitted and the charmm36.ff used in the conversion. > > I'd advise against using cgenff-derived parameters for serious > calculations, the charge penalties are usually pretty high and that means > you won't have a realistic description of your system. However, if you > really want to run a simulation using those parameters, that's perfectly > possible. That command generates a usable topology and coordinate pair, as > well as a parameter file that's loaded in the topology. For all intents and > purposes, using those for a single molecule simulation will work and should > be relatively straightforward. If you want to simulate a multi-molecule > system, you can either manually create a coordinate file modifying the > generated one or, alternatively, generate it using any molecule editing > software, exporting a pdb and using pdb2gmx to generate a > gromacs-compatible coordinate and topology file. > > > > *Joao Martins* > > joaomartins...@gmail.com > > On Wed, Nov 5, 2014 at 10:31 AM, Venkat Reddy wrote: > > > Dear all, > > I have a new lipid molecule and I want to simulate it using charmm36 > > forcefield. I have generated .str file using *paramchem* server and then > > using the following command, I got the itp file. > > > > ./cgenff_charmm2gmx.py COL COL.mol2 COL.str charmm36.ff > > > > Is this a reasonable approach for generating gromacs topology for a lipid > > molecule? If I change the atom types in the generated itp file to lipid > > atom types described in lipids.rtp, is it appropriate? > > > > Thank you for your time > > > > -- > > With Best Wishes > > Venkat Reddy Chirasani > > PhD student > > Laboratory of Computational Biophysics > > Department of Biotechnology > > IIT Madras > > Chennai > > INDIA-600036 > > -- > > Gromacs Users mailing list > > > > * Please search the archive at > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before > > posting! > > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > > > * For (un)subscribe requests visit > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > > send a mail to gmx-users-requ...@gromacs.org. > > > -- > Gromacs Users mailing list > > * Please search the archive at > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before > posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charmm to gromacs conversion for lipids
The rtp file is used by pdb2gmx for creating GROMACS topology and coordinate files. Changing it in the lipids.rtp will not do much other than allow you to generate from a pdb the .gro and .top for that molecule. Changing atom types is tricky and should be avoided in order to make sure you're using the cgenff-generated parameters, those atomtypes are connected to the molecule you submitted and the charmm36.ff used in the conversion. I'd advise against using cgenff-derived parameters for serious calculations, the charge penalties are usually pretty high and that means you won't have a realistic description of your system. However, if you really want to run a simulation using those parameters, that's perfectly possible. That command generates a usable topology and coordinate pair, as well as a parameter file that's loaded in the topology. For all intents and purposes, using those for a single molecule simulation will work and should be relatively straightforward. If you want to simulate a multi-molecule system, you can either manually create a coordinate file modifying the generated one or, alternatively, generate it using any molecule editing software, exporting a pdb and using pdb2gmx to generate a gromacs-compatible coordinate and topology file. *Joao Martins* joaomartins...@gmail.com On Wed, Nov 5, 2014 at 10:31 AM, Venkat Reddy wrote: > Dear all, > I have a new lipid molecule and I want to simulate it using charmm36 > forcefield. I have generated .str file using *paramchem* server and then > using the following command, I got the itp file. > > ./cgenff_charmm2gmx.py COL COL.mol2 COL.str charmm36.ff > > Is this a reasonable approach for generating gromacs topology for a lipid > molecule? If I change the atom types in the generated itp file to lipid > atom types described in lipids.rtp, is it appropriate? > > Thank you for your time > > -- > With Best Wishes > Venkat Reddy Chirasani > PhD student > Laboratory of Computational Biophysics > Department of Biotechnology > IIT Madras > Chennai > INDIA-600036 > -- > Gromacs Users mailing list > > * Please search the archive at > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before > posting! > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] Charmm to gromacs conversion for lipids
Dear all, I have a new lipid molecule and I want to simulate it using charmm36 forcefield. I have generated .str file using *paramchem* server and then using the following command, I got the itp file. ./cgenff_charmm2gmx.py COL COL.mol2 COL.str charmm36.ff Is this a reasonable approach for generating gromacs topology for a lipid molecule? If I change the atom types in the generated itp file to lipid atom types described in lipids.rtp, is it appropriate? Thank you for your time -- With Best Wishes Venkat Reddy Chirasani PhD student Laboratory of Computational Biophysics Department of Biotechnology IIT Madras Chennai INDIA-600036 -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
