cameronabrams
commented
6 months ago Hi Griffin,
Thanks for bringing this to my attention; I've never tried something like this. It looks to me like gromacs requires specific water parameters depending upon whether holonomic bondlength restraints are being used (the settle algorithm). If you modified the top file so that the #-lines are removed including the entire # else branch with the [settles] stanza, this would fix the problem, but then gromacs would run with flexible water (this is not bad, just inefficient). Doing the complementary thing will not work since the topology parser in HTPolyNet does not know about [settles] stanzas.
Your idea of using gmx solvate is good; it is out of the workflow of HTPolyNet, but should be ok. You could do a build through densification setting the pressure during densification to something like 10^-5, and then gmx solvate the output of that, and then do your own equilibrations after that.
I'll add this problem to the current buglist/future features list and hopefully I can find some time to get around to this soon; let me know if the gmx solvate thing works.
Cheers, Cameron
On Mon, May 20, 2024 at 4:19 PM Griffin Golde @.***> wrote:
First off, thank you for all the effort you put in creating and maintaining this package. I am using HTPolyNet to investigate the properties of poly(acrylamide-co-N,N'-methylenebisacrylamide) hydrogel. I have successfully generated the crosslinked polymer with no water present in the system, but since I am investigating swelling behavior, I am trying to simulate the polymerization in solution.
I have tried adding water directly to the constituents directive in the .yaml config file as if it were a monomer with no reactions, but receive the error:
File "/home/griffin/anaconda3/envs/mol-env/lib/python3.12/site-packages/HTPolyNet/topology.py", line 74, in typedata return int(s) ^^^^^^ ValueError: invalid literal for int() with base 10: '#ifdef'
I believe the error is happening when HTPolyNet is reading back in the H2O.top file it generates in proj-*/molecules/parameterized/ with the Topology class method read_top(). The generated H2O.top file contains the line "#ifdef FLEXIBLE" under the "[ atoms ]" directive that read_top() tries to convert to type int because it thinks the line is a row entry in under the [ atoms ] directive. Would modifying line 193 in Topology.py work to avoid this error (shown below)? Would it cause other errors?
Change line 193 in Topology.py from: "contentlines=[l for l in s[1:] if not (len(l.strip())==0 or l.strip().startswith(';'))]" to something like: "contentlines=[l for l in s[1:] if not (len(l.strip())==0 or ( l.strip().startswith(';') or l.strip().startswith('#') ) )]"
I am still an undergraduate and inexperienced in GROMACS. Are there any other issues you foresee by trying to simulate polymerization in solution? Would it be better to use "gmx solvate" after densification and before CURE? Thank you again for all your hard work.
My log files as well as my input H2O.mol2 file and the parameterized H2O.top file are attached (as .txt). console.log https://github.com/AbramsGroup/HTPolyNet/files/15381450/console.log diagnostics.log https://github.com/AbramsGroup/HTPolyNet/files/15381451/diagnostics.log H2O_mol2.txt https://github.com/AbramsGroup/HTPolyNet/files/15381452/H2O_mol2.txt H2O_top.txt https://github.com/AbramsGroup/HTPolyNet/files/15381453/H2O_top.txt
— Reply to this email directly, view it on GitHub https://github.com/AbramsGroup/HTPolyNet/issues/4, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEFHYWXIXLTIY5K7CWJIQ2TZDJLD3AVCNFSM6AAAAABIAKIBK6VHI2DSMVQWIX3LMV43ASLTON2WKOZSGMYDMNZSGIYTAOA . You are receiving this because you are subscribed to this thread.Message ID: @.***>
ggolde
First off, thank you for all the effort you put in creating and maintaining this package. I am using HTPolyNet to investigate the properties of poly(acrylamide-co-N,N'-methylenebisacrylamide) hydrogel. I have successfully generated the crosslinked polymer with no water present in the system, but since I am investigating swelling behavior, I am trying to simulate the polymerization in solution.
I have tried adding water directly to the constituents directive in the .yaml config file as if it were a monomer with no reactions, but receive the error:
File "/home/griffin/anaconda3/envs/mol-env/lib/python3.12/site-packages/HTPolyNet/topology.py", line 74, in typedata return int(s) ^^^^^^ ValueError: invalid literal for int() with base 10: '#ifdef'
I believe the error is happening when HTPolyNet is reading back in the H2O.top file it generates in proj-*/molecules/parameterized/ with the Topology class method read_top(). The generated H2O.top file contains the line "#ifdef FLEXIBLE" under the "[ atoms ]" directive that read_top() tries to convert to type int because it thinks the line is a row entry in under the [ atoms ] directive. Would modifying line 193 in Topology.py work to avoid this error (shown below)? Would it cause other errors?
Change line 193 in Topology.py from: "
contentlines=[l for l in s[1:] if not (len(l.strip())==0 or l.strip().startswith(';'))]" to something like: "contentlines=[l for l in s[1:] if not (len(l.strip())==0 or ( l.strip().startswith(';') or l.strip().startswith('#') ) )]"I am still an undergraduate and inexperienced in GROMACS. Are there any other issues you foresee by trying to simulate polymerization in solution? Would it be better to use "gmx solvate" after densification and before CURE? Thank you again for all your hard work.
My log files as well as my input H2O.mol2 file and the parameterized H2O.top file are attached (as .txt). console.log diagnostics.log H2O_mol2.txt H2O_top.txt