Hi Joe,
This is caused by errors in the MPI messages: perhaps network failures or
bugs on the network cards.
However, as long as you have the files you mention, you're lucky and this
is rescuable. :-)
Use relion_reconstruct -- debug --debug_ori_size 200 --debug_size 200 --i
run1_half1
(where run1 is the rootname of your run, and 200 is the size of your
images. Do this also for half2. Save the resulting maps with the same
unfil.mrc map as you would normally get and then proceed with
postprocessing as usual.
HTH,S
> Hi all,
>
> I used relion-1.3 to carry out the structure determination of my protein.
> However, at the end of auto-refine, the job stopped. Instead of getting
> the 3D models from half1 or half2, I got these files:
>
> _half2_class001_data_real.mrc
> _half1_class001_data_real.mrc
> _half2_class001_data_imag.mrc
> _half1_class001_data_imag.mrc
> _half2_class001_weight.mrc
> _half1_class001_weight.mrc
>
> The error code is like:
> MPI_ABORT was invoked on rank 2 in communicator MPI_COMM_WORLD with
> errorcode 1.
>
> I tried to run this on either a cluster or a multi-core desktop, it always
> stopped at the same point.
>
> Any suggestion will be much appreciated.
>
> Kind regards,
> Joe
>
--
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres
|