From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Wed Apr 29 2020 - 01:59:05 CDT
Hi Pawel
I am QMMM investigating three small-organic-molecule inhibitors of a
nucleoside system that only differ in the nature of the halogen.
*TmpOut lists Mulliken charges for C-halo as follows
C(0.08)-I(-0.21) "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
"%%basis NewGTO I \"old-ZORA-TZVP\" end end"
(QMMM > ten-fold slower that the two below)
C(-0.23)-Br(0.29) "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
C(-0.01)-Cl(0.10) "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
Although probably the iodine atom can't be seen as definitely anionic and
the pyrrole carbon, to which it is attached, cationic, the trend is
definitely what you expected.
I overlooked, me bad, the problem of the incompleteness of the basis set
and I thank you for your suggestions.
It would be difficult for me to set the simulation in order for the iodine
case, and, at any event, the slowness of the simulation would pose great
problems. The best core ratio that I found for NAMD/ORCA was 1/34, with all
the consequences.
cheers
francesco
On Tue, Apr 28, 2020 at 10:10 PM Pawel Kedzierski <
pawel.kedzierski_at_pwr.edu.pl> wrote:
> W dniu 28.04.2020 o 19:06, Francesco Pietra pisze:
>
> Hi Mariano
> the QM-MM (NAMD-ORCA) is running well with commands passed from NAMD to
> ORCA
>
> qmConfigLine "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
>> qmConfigLine "%%basis NewGTO I \"old-ZORA-TZVP\" end end"
>> qmConfigLine "%%maxcore 2500"
>> qmConfigLine "%%pal nproc 34 end"
>> qmConfigLine "%%output Printlevel Mini Print\[ P_Mulliken \] 1
>> Print\[P_AtCharges_M\] 1 end"
>> qmConfigLine "%%scf SOSCFStart 0.00033 end" # Default value of orbital
>> gradient is 0.0033
>>
>
> However, to my great surprise, the C-I bond length has increased from the
> initial 2A (NAMD MD) to 2.5A, and now to 3A with QMMM. Do you have any idea
> why ORCA is computing the C-I so much wrongly? That part of the structure
> is an iodopyrrole, i.e., the I-atom is at at an sp2 C in a five-membered
> ring.
>
> I do not think this is a bug in ORCA, which is following your
> instructions. My guess is that you introduced the problem yourself by
> adding a triple zeta basis set for iodine, while the rest of the system has
> SVP.
>
> This is one of many possible artifacts introduced by incompleteness of the
> basis set. If the basis set is neither close to complete (which is too
> expensive even for most of single point calculations, save QMMM), nor is it
> balanced, the electrons flow where there are more basis set functions
> available. In your case, I think your iodine attracted electrons so much
> that it became an anion and there is no covalent bond any more between it
> and the (now cationic) pyrrole. You should be able to find the ORCA output
> files in the temporary folder given by qmBaseDir NAMD variable; check the
> atomic charges and the Mayer bond orders to see if I'm right.
>
> I would advice using a consistent basis set for all QM atoms, like
> LANL2DZ, or maybe MINIX if You wan a minimal one, but it may be some
> challenge to find a basis set with iodine and a matching auxiliary basis
> for RIJCOSX.
>
> HTH,
>
> Pawel
>
>
> Thanks for advice
> francesco
>
>
>
> On Thu, Apr 16, 2020 at 7:01 PM Mariano Spivak <marspivak_at_gmail.com>
> wrote:
>
>> Yes, my bad.
>> For that you need to write it as:
>> qmConfigLine "%%basis NewGTO I \"old-ZORA-TZVP\" end end"
>>
>>
>> On Thu, Apr 16, 2020 at 11:52 AM Francesco Pietra <chiendarret_at_gmail.com>
>> wrote:
>>
>>>
>>> Hi Mariano
>>> With code
>>>
>>> qmConfigLine "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
>>>> qmConfigLine "%%basis NewGTO I "old-ZORA-TZVP" end end"
>>>> qmConfigLine "%%maxcore 2500"
>>>> qmConfigLine "%%pal nproc 34 end"
>>>> qmConfigLine "%%output Printlevel Mini Print\[ P_Mulliken \] 1
>>>> Print\[P_AtCharges_M\] 1 end"
>>>> qmConfigLine "%%scf SOSCFStart 0.00033 end" # Default value of
>>>> orbital gradient is 0.0033
>>>>
>>>
>>> namd log complains
>>>
>>> TCL: extra characters after close-quote
>>>
>>> FATAL ERROR: extra characters after close-quote
>>> while executing
>>> "qmConfigLine "%%basis NewGTO I "old-ZORA-TZVP" end end"
>>> qmConfigLine "%%maxcore 2500"
>>> qmConfigLine "%%pal nproc 34 end"
>>> qmConfigLine "%%o..."
>>> (file "namd-01.conf" line 173)
>>> [Partition 0][Node 0] End of program
>>>
>>>
>>> This is the same type of error that I encountered initially. The basis
>>> set needs to enclosed within " " but the whole line is incorrect.
>>>
>>> Best
>>> francesco
>>>
>>> On Thu, Apr 16, 2020 at 6:24 PM Mariano Spivak <marspivak_at_gmail.com>
>>> wrote:
>>>
>>>> Francesco,
>>>>
>>>> In that case, you can keep the line
>>>> qmConfigLine "%%basis NewGTO I "old-ZORA-TZVP" end end"
>>>> but define the basis set for the non-iodine atoms in the first line,
>>>> for example:
>>>> qmConfigLine "! B3LYP D3BJ SVP RIJCOSX enGrad Opt SlowConv"
>>>>
>>>> Best
>>>>
>>>>
>>>> On Thu, Apr 16, 2020 at 10:42 AM Francesco Pietra <
>>>> chiendarret_at_gmail.com> wrote:
>>>>
>>>>> Hi Mariano
>>>>> That clarifies the matter. However, going to def2-TZVP basis set for
>>>>> all atoms would be too costly, probably unnecessary, and surely prevented
>>>>> on the single node I am using.
>>>>>
>>>>> Could you please set the code in order for having an appropriately
>>>>> high basis set for the iodine-atom only?
>>>>>
>>>>> Thanks
>>>>> francesco
>>>>>
>>>>> On Thu, Apr 16, 2020 at 5:33 PM Mariano Spivak <marspivak_at_gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Francesco,
>>>>>>
>>>>>> The first qmConfigLine is not quite right. You have the SV keyword
>>>>>> which supersedes the def2-TZVP basis set, giving you error since the SV
>>>>>> refers to the minimal basis set which is not available for Iodine. If you
>>>>>> remove that it should work. You don't need the NewGTO part.
>>>>>>
>>>>>> Thanks
>>>>>>
>>>>>> On Thu, Apr 16, 2020 at 10:22 AM Francesco Pietra <
>>>>>> chiendarret_at_gmail.com> wrote:
>>>>>>
>>>>>>> I forgot to mention that without the NewGTO line, when atom-I
>>>>>>> preset, reported
>>>>>>>
>>>>>>> /0/*TmpOut:
>>>>>>> ----- Orbital basis set information -----
>>>>>>> Your calculation utilizes the basis: SV
>>>>>>>
>>>>>>> ----- AuxJ basis set information -----
>>>>>>> Your calculation utilizes the auxiliary basis: def2/J
>>>>>>> F. Weigend, Phys. Chem. Chem. Phys. 8, 1057 (2006).
>>>>>>>
>>>>>>>
>>>>>>> *****************************************************************
>>>>>>> ** There are no main basis functions on atom number 241 (I
>>>>>>> ) **
>>>>>>>
>>>>>>> *****************************************************************
>>>>>>> [file orca_main/maininp2.cpp, line 2348]: The basis set was
>>>>>>> either
>>>>>>> not assigned or not available for this element - Aborting the run
>>>>>>> ---------- Forwarded message ---------
>>>>>>> From: Francesco Pietra <chiendarret_at_gmail.com>
>>>>>>> Date: Thu, Apr 16, 2020 at 5:11 PM
>>>>>>> Subject: namd-l: Error: Number of NGauss expected in AddNewGTOs
>>>>>>> To: <GerardR_at_usca.edu>, NAMD <namd-l_at_ks.uiuc.edu>
>>>>>>>
>>>>>>>
>>>>>>> Hi Gerard
>>>>>>> With QMMM, I am trying to enhance the basis set for only the iodine
>>>>>>> atom substituent at carbon of a small organic ligand.
>>>>>>>
>>>>>>> Code (OK without NewGTO for not very heavy atoms)
>>>>>>> qmConfigLine "! B3LYP D3BJ def2-TZVP SV RIJCOSX enGrad Opt
>>>>>>> SlowConv"
>>>>>>> qmConfigLine "%%basis NewGTO I "old-ZORA-TZVP" end end"
>>>>>>> qmConfigLine "%%maxcore 2500"
>>>>>>> qmConfigLine "%%pal nproc 34 end"
>>>>>>> qmConfigLine "%%output Printlevel Mini Print\[ P_Mulliken \] 1
>>>>>>> Print\[P_AtCharges_M\] 1 end"
>>>>>>> qmConfigLine "%%scf SOSCFStart 0.00033 end" # Default value of
>>>>>>> orbital gradient is 0.0033
>>>>>>>
>>>>>>> ERROR: Number of NGauss expected in AddNewGTOs
>>>>>>>
>>>>>>> I have seen a post of last year about such error message but, as
>>>>>>> reported, it remained unsolved.
>>>>>>> Here probably the grid should also be enhanced.
>>>>>>>
>>>>>>> All the best
>>>>>>> francesco pietra
>>>>>>>
>>>>>>
>
This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:08 CST