From: Eduardo R Cruz-Chu (cruzchu_at_uwm.edu)
Date: Fri Nov 04 2022 - 14:12:44 CDT
Hi Chunli;
I used namd for about 6 million atoms about a decade ago. As I recall, the key problem was that the standard PDB/PSF formats were not designed to store so much data, and run out of digits. The solution was to use the JS format. More info in :
https://www.ks.uiuc.edu/Research/namd/wiki/index.cgi?NamdMemoryReduction
As the PDB format has a fixed number of digits to store positions, avoid that and better use COOR files (bincoor).
Also, for genereting your structures, stick with PSFGEN. If there were an error in PSFGEN, then fix it at that step; otherwise that error will pop up again during visualization/simulation.
Regards;
Eduardo
On Nov 4, 2022, at 12:57 PM, Bunzeki Kilolo (Mme) <kilolo.bunzeki_at_polytechnique.edu<mailto:kilolo.bunzeki_at_polytechnique.edu>> wrote:
Dear Chunli,
I suggest you to create a smaller system around the interaction site that you want to study.
Best,
Kilolo
________________________________
De: "Chunli Yan" <utchunliyan_at_gmail.com<mailto:utchunliyan_at_gmail.com>>
À: "NAMD list" <namd-l_at_ks.uiuc.edu<mailto:namd-l_at_ks.uiuc.edu>>
Envoyé: Vendredi 4 Novembre 2022 18:44:22
Objet: namd-l: NAMD can not run 5.6 million system with amber prmtop
Hello,
I have a protein-DNA system with a total atom number ~5.6 million. After we generated the amber prmtop files, we can not use VMD to visualize them (in vmd: PARM7: error reading bond number 3995983) and also can not run with NAMD.
I am wondering whether the system is too big and can not run (we do not have issues for smaller systems with amber prmtop), whether charmm psf will work for larger systems.
Best,
Chunli
This archive was generated by hypermail 2.1.6 : Tue Dec 13 2022 - 14:32:44 CST