Memory limitations with all-atom NMA

[Former user]

Hi, I'm a relatively new user of Bio3D, and I have noticed that when I do all-atom NMA for large structures (with RTB), the upper limit is something like 2000 residues, above that aanma crashes. It seems to be a memory-related thing, but increasing the available memory does not help beyond a certain point, e.g. 64GB or 128GB results in little or no difference, the upper limit on my system remains roughly 2000 residues (with 64GB). I would like to ask whether there is some way around this, or one just has to use coarse grained (e.g. Ca) nma.

Another (minor) thing, I am assuming that in the "dccm" function the "nmodes" argument uses only the non-trivial modes, (so nmodes=3 are actually modes 7,8 and 9); am I correct with this?

Kind regards, George Abrusan

[Xinqiu Yao]

Hi,

‌

For the first question, I think currently there is no good solution except using the coarse-grained model or increasing computer memory. One possible future direction may be to use the ‘big.memory’ package for large systems.

‌

For the second, yes, it means the number of ‘non-trivial’ modes.

[harold hilarion Fokoue]

Hi, please how do we used “bigmemory” for large system?