[gmx-users] Pull Code problems

[DimitryASuplatov]

OK,
I have a new problem.
I`ve started the simulation with the following pull parameters


;-----------PULL CODE ---------------------

pull<-->=<----->  umbrella
pull_geometry<->= direction
pull_group0<-->= r_500
pull_group1<--->= r_535
pull_vec1<----->= -1.306 1.355 -0.319
pull_k1><------>= 1000
pull_rate1<---->= 1 ; nm/ps = 1 A/ns

With such extreme pulling velocities I had my system teared apart in 100
ps - that is OK I suggest. The problem is that pull group (pull_group1)
and reference group (pull_group0) were both pulled in opposite
directions - pull_vec1 and -(pull_vec1)?!

1/ In my best understanding the reference group is the one that stays
intact to track the movement of the group that is being pulled - pull
group. Then how could this happen in my system? What should I do to pull
only the pull_group1?
2/ I was recomended to use AFM pull mode in gromacs 3. What is the new
name for AFM in gromacs 4? 

Thanks.