What causes 'Segmentation fault' during well-tempered metadynamics?

[Seulwoo Kim]

Hi all,

I've struggled with the 'Segmentation fault' running metadynamics with PLUMED 2.3.0.

What I'm trying to do is calculation of free energy of a molecule varing with altitude angle and azimuthal angle, so I used well-tempered metadynamics technique.

All the processes of calculating angles are done with MATHEVAL command, and as the molecule is symmetrical, I adjust the altitude angle ranged 0 to 90 degree.

I catched 2 kinds of error:

1. LINCS warning -> segentation fault (not majority)

2. Segmentation fault (solely, majority)

And these happened only I used METAD command option.

I tried so many trials to fix this errors but not effective..

Here are some trials I did

1. Adjust all the parameters of METAD : GRID_BIN, SIGMA, HEIGHT, GAMMA, whether GRID_SPARSE ON/OFF, ...

2. Give restrains on torsion of dihedral angles and distance between surface and the molecule on z direction : Because what I want is not the effect of dihedral but the geometrical angles. Restrain on distance is for the crystallization.

3. Start with varied initial configurations

I wonder that, "WHAT causes 'Segmentation fault'" and HOW can I overcome this situation?!?!

I just post the PLUMED .dat file and error log.

Thank you for reading and please your kind answers !

Have a nice day !

Seulwoo Kim

[Giovanni Bussi]

Hi,

there's a way easier manner to compute these angles:

theta: check the first example in https://plumed.github.io/doc-v2.3/user-doc/html/_f_i_x_e_d_a_t_o_m.html

phi: you can build an equivalent construction using a TORSION

Still, notice that when theta is zero phi is discontinuous, and adding forces to it might lead to crashes.

It might be better to project the vector on a sphere of unitary radius and then as three coordinates the three components

Giovanni

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/ee867f38-4e19-4c23-979c-9982636eaa5c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Seulwoo Kim]

Thanks to reply, Giovanni !!

And I'm sorry about giving late my thanks because I was a little busy :(

I ran my metad simulation following your techniq, set CVs as three coordinates x y and z.

Fortunatly, I finally got the message gmx gives the comment which means the simulation complete !

Still, I have something to struggle in post-processing step..

"Does any method exist that summate each bias as following theta and phi angle?I mean transformation x y z result into theta phi result."

I firstly tried to handle my result from directly derived from sum_hills and convert them into the angles.

Then, I faced the problem that the interval is not consistent and the dimension was not matched each other.

So I want to make these grid following angle have equal interval.

Thanks,

Seulwoo

2017년 9월 13일 수요일 오후 9시 56분 58초 UTC+9, Giovanni Bussi 님의 말:

To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.