Re: [AMBER] ValueError: could not convert string to float: ********* Error occured on rank 0. Exiting. All files have been retained.

From: David Case <>
Date: Thu, 16 Jan 2020 13:36:00 +0000

On Thu, Jan 16, 2020, Neha Gandhi wrote:
>I am running mmgbsa and single residue decomposition. It threw following
>error "ValueError: could not convert string to float: *********

Look at your trajectory file with a text editor: you will see that there
are coordinates that are too large to fit into the space allowed, and
are indicated by stars.

Simplest thing usually is to set ioutfm=1, and write to a netCDF
trajectory format. However, there may also be something wrong with your
system, and it is "blowing up" somehow. Be sure to visualize things
with a program like chimera or VMD or pymol, to help make sure that you
are not sending bad results to MM_PBSA.

Note also the opinion of many: for snapshot analysis (like MM_PB/GBSA),
the PB model is generally more robust than in GB. GB can be very useful
for running trajectories, since it is fast, and can be applied at every
step. But where speed is not needed so much, PB can be very helpful.
So, at the least, consider running MM_PBSA alongside MM_GBSA.


AMBER mailing list
Received on Thu Jan 16 2020 - 06:00:04 PST
Custom Search