Sasha,
thanks for the update, and yes, you're right that all parameter
information is available.
However, your input might contain a formatting error in your two lines
for the torsion with the central atoms C*-S:
Since the atom types contain two characters like 'CT' or 'C*', a space
character is used and required in cases where the atom type name is just
a single character like 'S ' or 'C '.
From the documentation
(
https://ambermd.org/FileFormats.php#frcmod) it can be deduced that leap
expects the format of a torsion to follow a strict format, i.e. four
two-character strings separated by a single position.
Hopefully, inserting two spaces at the correct positions will make your
frcmod file valid. Please try
2C-S-C*-C* -> 2C-S -C*-C*
2C-S-C*-C -> 2C-S -C*-C
(When you successfully created a valid parameter file, please do not
forget to run some validation calculations to test the parameter set and
maybe update/change it. I mention that explicetly, because the process
of creating a technically valid is often that cumbersome that one might
easily want to forget the second step and run simulations, finally.)
Best,
Anselm
Bioinformatik | NHR.FAU
Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU)
Germany
On 02/28/2024 12:49 AM, Sasha Buzko wrote:
> Anselm,
> Thank you for your response.
> I did check the frcmod and admittedly I should have included the relevant fragments in my previous post. I'm including these lines where I copied the atom type, bond, angle and torsion parameters (all for consistency, not sure whether necessary). The upper line is the original generated for the CYX-linker structure.
>
> CT 12.010 0.878 same as c3
> 2C 12.010 0.878 same as c3
>
> ...
> CT-S 215.90 1.839 same as c3-ss, penalty score= 0.0
> 2C-S 215.90 1.839 same as c3-ss, penalty score= 0.0
>
> ...
> C*-S -CT 61.100 100.640 same as c3-ss-cc, penalty score= 0.0
> C*-S -2C 61.100 100.640 same as c3-ss-cc, penalty score= 0.0
>
> ...
> C*-C*-S -CT 1 1.100 180.000 -2.000 same as c2-c2-ss-c3
> 2C-S-C*-C* 1 1.100 180.000 -2.000 same as c2-c2-ss-c3
> C -C*-S -CT 1 1.100 180.000 -2.000 same as c2-c2-ss-c3
> 2C-S-C*-C 1 1.100 180.000 -2.000 same as c2-c2-ss-c3
>
>
> So in theory, the loaded frcmod contains the necessary parameters (unless something strange is visible, of course). The corresponding prepin has the atom type updated to 2C as well. And when tleap loads the structure and I run check, it returns no errors. It only complaints at the final stage of writing out the prmtop and incprd.
>
> I'm baffled here and any hints would be much appreciated.
>
> Thank you!
>
> Sasha
>
>
>
> ---
> Oleksandr “Sasha” Buzko
> Computational Biology
> ImmunityBio
> 9922 Jefferson Blvd
> Culver City, CA 90230
>
>
> -----Original Message-----
> From: Dr. Anselm Horn <anselm.horn.fau.de>
> Sent: Tuesday, February 27, 2024 11:31 AM
> To: Sasha Buzko <Sasha.Buzko.ImmunityBio.com>; David A Case <dacase1.gmail.com>; AMBER Mailing List <amber.ambermd.org>
> Subject: Re: [AMBER] Is it possible to model a protein with a linker covalently attached at a cysteine? Missing angle parameters
>
> This email originated from outside of NANT.
> Do not click links or open attachments unless you recognize the sender and are expecting the content.
> If you believe this email is suspicious, report using “Report Phish” button.
>
>
>
> Sasha,
>
> since you are already far in your parameterization efforts, it's not that easy to figure out your problem without being able to inspect the files.
> For parameterization, you built a system with a CYX residue and your linker, and successfully extracted the linker-related information, i.e.
> you have a prepin (chemical/topological definition with atom types and
> charges) and a frcmod (parameters) file of your linker residue. Ok, so your approach is to re-use CYX in your system and provide a linker residue "from the other side", right?
>
> The leap error tells you that the interface between the CYX and the new linker residue is not fully defined, i.e. your frcmod file lacks torsional parameters for the X-S-C*-X torsion, more specifically the two torsions explicitely printed.
>
> I now guess you intend to use the Amber forcefield parm14 or parm19, which has the CB defined as the atom type 2C. (The automatic atom type assingment during your parameterization yielded a CT, and you generated a frcmod file with that atom type; but according to your e-mail you changed that already.)
>
> To confirm that the linkage is the problematic part, omit the bond command between the CYX and your linker in leap and try to save a top/crd file pair.
>
> I further guess that the problem is somewhere in the frcmod file:
> Have a look into it to see, whether it describes the two torsions leap is complaining about:
> 2C-S-C*-C*
> 2C-S-C*-C
>
> Since both do not occur in parm14/parm19, the must be provided in the frcmod file.
> Check the atom types in your prep file, and add a "general" torsonal term for the linkage to the frcmod file like X -S-C*-X to see whether this gives you a top/crd pair.
>
> Again, without the actual files, it's hard to give a specific suggestion.
>
> Good luck,
>
> Anselm
>
> Bioinformatik | NHR.FAU
> Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Germany
>
>
>
>
>
>
> Am 27.02.2024 um 05:31 schrieb Sasha Buzko:
>> David, Anselm,
>> Thank you for much for the advice.
>> I've built a linker with the cystine bound to it and ran it through antechamber to calculate the parameters and then generated the corresponding frcmod. After some iterations I've managed to get tleap to read in all necessary values and stop complaining (check model comes back clean). However, I'm hitting one (hopefully last) speedbump:
>>> saveamberparm x a.prmtop a.inpcrd
>> Checking Unit.
>>
>> Note: Ignoring the warnings from Unit Checking.
>>
>> Building topology.
>> Building atom parameters.
>> Building bond parameters.
>> Building angle parameters.
>> Building proper torsion parameters.
>>
>> Error: ** No torsion terms for 2C-S-C*-C*
>>
>> Error: ** No torsion terms for 2C-S-C*-C Building improper torsion
>> parameters.
>> old PREP-specified impropers:
>> <LAX 579>: C8 C6 N2 H15
>> <LAX 579>: C7 N2 C8 O2
>> <LAX 579>: C11 C16 N4 H20
>> <LAX 579>: C18 N4 C11 O5
>> <LAX 579>: C12 C15 N5 C19
>> <LAX 579>: C13 N5 C12 O6
>> <LAX 579>: C14 C12 C13 H30
>> <LAX 579>: C14 N5 C15 O7
>> <LAX 579>: C1 N3 C2 O1
>> <LAX 579>: C2 H26 N3 H27
>> total 1796 improper torsions applied
>> 10 improper torsions in old prep form Building H-Bond parameters.
>> Incorporating Non-Bonded adjustments.
>>
>> Warning: Parameter file was not saved.
>>
>>
>> The frcmod.lax file actually has the right torsion parameters but the atom type of the CB in cystine connected to the linker was assigned by antechamber as CT and apparently tleap was expecting 2C as defined for CYX. I manually changed the atom type in the frcmod and added corresponding lines under the torsions, angles, bonds and atom types used. I've also changed the atom type in the prepin file to match it in the CYX, but still get the same error.
>>
>> Do you know why check didn't complain but the final structure saving step found a problem? Also, is there a reasonable way to resolve this conundrum? Am I actually doing it right (more or less) or am I missing something obvious?
>>
>> Thanks again!
>>
>> Sasha
>>
>>
>> ---
>> Oleksandr “Sasha” Buzko
>> Computational Biology
>> ImmunityBio
>> 9922 Jefferson Blvd
>> Culver City, CA 90230
>>
>>
>> -----Original Message-----
>> From: David A Case <dacase1.gmail.com>
>> Sent: Monday, February 26, 2024 5:59 PM
>> To: Sasha Buzko <Sasha.Buzko.ImmunityBio.com>; AMBER Mailing List
>> <amber.ambermd.org>
>> Cc: Dr. Anselm Horn <anselm.horn.fau.de>
>> Subject: Re: [AMBER] Is it possible to model a protein with a linker
>> covalently attached at a cysteine? Missing angle parameters
>>
>> This email originated from outside of NANT.
>> Do not click links or open attachments unless you recognize the sender and are expecting the content.
>> If you believe this email is suspicious, report using ?Report Phish? button.
>>
>>
>>
>> On Mon, Feb 26, 2024, Sasha Buzko via AMBER wrote:
>>
>>> Thank you for your suggestion to use CYX as the branching point. I
>>> have the parameters for the linkers (from antechamber) and tleap is
>>> not complaining about the branching. However, it gives an error
>>> message when checking the
>>> structure:
>>> Error: Could not find angle parameter: S - CT - C
>>>
>>> Error: Could not find angle parameter: S - CT - HC
>>>
>>> Error: Could not find angle parameter: S - CT - HC
>>>
>>> Looks like it's having trouble with the bonds around the sulfur in
>>> the CYX residue. I have loaded gaff and gaff2 but it still gives this
>>> error. Can you think of a way to get around this?
>>
>> Atom types using upper case letters (as above) are for Amber atom types.
>> GAFF used lower case letters for atom types, so loading those files won't do you any direct benefit. But you could look in gaff2.parm for similar angles (just changing upper to lower case): then make a frcmod file with the upper-case equivalents.
>>
>> Or, you may just wish to change the atom types above. Methionine in Amber force fields has S-CT-H1 (instead of S-CT-HC). I don't know the chemistry involved in your linker, but amino12.lib (used for ff19SB) has 2C-2C-S, which may be close to what you need.
>>
>> The "general" problem is that antechamber is mainly used for non-covalent ligands. If you want to link to an amino acid, you probably want (as you are doing) to use Amber atom types. But then you need to be inventive to find missing angle parameters. You can generally check what you decide upon by minimizing the system, and looking carefully at the resulting geometry around you new chemistry.
>>
>> ...hope this helps...dac
>>
>> p.s. sorry to insert myself into this conversation. Dr. Horn may have more specific advice here.
>>
>> CONFIDENTIALITY NOTICE
>> This e-mail message and any attachments are only for the use of the intended recipient and may contain information that is privileged, confidential or exempt from disclosure under applicable law. If you are not the intended recipient, any disclosure, distribution or other use of this e-mail message or attachments is prohibited. If you have received this e-mail message in error, please delete and notify the sender immediately. Thank you.
>>
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 28 2024 - 00:30:04 PST
