Linked-cell - static protein + buffer gas:
I have a simulation involving a static protein hit by a gas (He for example). I want to calculate the LJ potential between the gas and the protein atoms using the linked-cell algorithm. My problem is that the neighbors are not being scanned correctly. Below the piece of my code where I am trying to identify the neighbors:
We have the gas (He) coordinates: mc[0] = x, mc[1] = y and mc[2] = z.
With this we can obtain the cell index for the gas:
/* Translate the vector cell index, mc, to a scalar cell index */
c = mc[0]*lcyz+mc[1]*lc[2]+mc[2];
And then scan the neighbor cells:
/* Skip this cell if empty */
if (head[c] == EMPTY) continue;
/* Scan the neighbor cells (including itself) of cell c */
for (mc1[0]=mc[0]-1; mc1[0]
A documented example of the linked-cell algorithm is in this document:
http://cacs.usc.edu/education/cs596/01-1LinkedListCell.pdf
---
This is not my area.
Dear Martin Klvana,
Thank you for your attention.
I am using exactly this algorithm to write my code. But all the algorithms I found were examples for a conventional MD simulation. My problem is quite different, because I have a static protein hit by a buffer gas. At the beginning of the simulation the gas are distant from the protein (no neighbor exist here), but when the gas are near to the protein I am not scanning correctly the neighbors.
I don't know but shouldn't it be actually easier to model a rigid protein than a dynamic one?
- Badges
- Science topic