[gmx-users] Re:Gromacs auxillary tools ideas

[lloyd riggs]

Dear All,

I read a paper by Van der Groot and realized some tool regarding simple paersons r values, or correlation co-efficient by different names would be a great analytical asset for macromolecular biology.  Looking over the code, g:sham seems to have begun implicating such things but it loos like a caotic mess.  Basically, I used gnumeric, but MS excel also has functions, and you can take say a PC analysis, and run any other data set and see if it "Correlates", ie is an effect of or not related in any way too some other factor, in the case of PC, energy versus anything else.  All they need is say you take your Gaussian 10-300 trajectories as a mean in a spread sheet (ie 200-5000) data points in a column, and compare it with PC (or any other thing you want, although energy is about the most comprehensive), and it gives a linear correlation based on pearsons two column correlation of relation (circa 1895)...it does allow say for a a PC, then taking several things like distance, hydro
 gen bonding means between 2 prtoeins, domain distance, inter protein movements, inter protein hydrogen bonding, LJ or coloumb, etc... and tells you which is the energy or correlation...would be a great asset...

And any competent programmer namless could do it in a week or two...as a gromacs tool...would add Bio more to funding sources...

Sincerely,

Stephan Watkins