CTFFIND4.1.8 Error

Submitted by iyerka on Tue, 2018-03-20 16:34

Forums

ctffind & ctftilt

Hello,

I am trying to use Relion-2.1-beta-1 and CTFFIND4.1.8 however, I keep getting the following error: "WARNING: skipping, since cannot get CTF values for Micrographs/mic000001.mrc" for every micrograph. An empty micrographs_ctf.star file is generated.

The note.txt file reads
"++++ Executing new job on Mon Mar 19 18:49:23 2018
++++ with the following command(s):
`which relion_run_ctffind_mpi` --i Import/job001/micrographs.star --o CtfFind/job002/ --CS 2.7 --HT 300 --AmpCnst 0.1 --XMAG 10000 --DStep 0.69 --Box 512 --ResMin 30 --ResMax 5 --dFMin 5000 --dFMax 70000 --FStep 500 --dAst 100 --use_noDW --ctffind_exe /gpfs_fs/cryoem/software-install/ctffind-4.1.8-exe/bin/ctffind --ctfWin -1 --is_ctffind4
++++".

The run.err file reads "WARNING: skipping, since cannot get CTF values for"

The run.out file reads
" === RELION MPI setup ===
+ Number of MPI processes = 1
+ Master (0) runs on host = godel35
=================
Using CTFFIND executable in: /gpfs_fs/cryoem/software-install/ctffind-4.1.8-exe/bin/ctffind
to estimate CTF parameters for the following micrographs:
*
*
*
*
Estimating CTF parameters using Niko Grigorieff's CTFFIND ...
Done! Written out: CtfFind/job002/micrographs_ctf.star"

The run.job reads
"job_type == 2
is_continue == false
Magnified pixel size (Angstrom): == 0.69
Nr of movie frames to average: == 4
FFT box size (pix): == 512
Spherical aberration (mm): == 2.7
Estimate CTF on window size (pix) == -1
Amount of astigmatism (A): == 100
Maximum defocus value (A): == 70000
Minimum defocus value (A): == 5000
Defocus step size (A): == 500
Perform equi-phase averaging? == No
Ignore 'Searches' parameters? == Yes
Estimate Thon rings from movies? == No
Estimate phase shifts? == No
Submit to queue? == Yes
CTFFIND-4.1 executable: == /gpfs_fs/cryoem/software-install/ctffind-4.1.8-exe/bin/ctffind
Gctf executable: == <98>µ~
Which GPUs to use: ==
Input micrographs STAR file: == Import/job001/micrographs.star
Voltage (kV): == 300
Minimum dedicated cores per node: == 1
Movie rootname plus extension == _movie.mrcs
Number of MPI procs: == 40
Additional arguments: ==
Other Gctf options: ==
Phase shift (deg) - Max: == 180
Phase shift (deg) - Min: == 0
Phase shift (deg) - Step: == 10
Amplitude contrast: == 0.1
Queue submit command: == qsub -q all.q@godel35
Standard submission script: == /gpfs_fs/cryoem/software-install/relion-v2-f-godel/bin/qsub-serial.bash
Queue name: == openmpi
Maximum resolution (A): == 5
Minimum resolution (A): == 30
Use CTFFIND-4.1? == Yes
Use Gctf instead? == No
Use micrograph without dose-weighting? == Yes"

The _ctffind4.log reads
"** Welcome to Ctffind **

Version : 4.1.8
Compiled : Jun 7 2017
Mode : Scripted

Input image file name :
Error: Blank answer in scripted mode, exiting..."

I am a novice user and any help would be appreciated.

Regards,
Kavita

Alexis Tue, 2018-03-20 19:24

Hi Kavita,

The relevant error message from ctffind is "Error: Blank answer". That, and the line above it, clearly show that your relion script did not submit the micrograph filename to ctffind and instead gave a blank line as input.

To convince yourself that ctffind is OK, you can just call the program directly from the command line and give it one of your micrographs as input.

I am not sure where the problem is in your Relion, but you should look carefully at the syntax you're using, compare it to the Relion tutorial, and if the .star file is one you came up with yourself, maybe double-check that also.

I hope this helps a little,
Alexis

iyerka Wed, 2018-03-21 11:12

Hi Alexis,

Thank you for getting back so quickly.

I tried running the Relion2 tutorial. I copied the Import and Motioncorr folders from the precalculated results. I tried running CTFFIND4.1 on "betagal/MotionCorr/job002/corrected_micrographs.star" and got the same result.

run.err ->
"WARNING: skipping, since cannot get CTF values for MotionCorr/job002/Micrographs/Falcon_2012_06_12-14_33_35_0_noDW.mrc" .... for every micrograph

And the individual log files say the following ->
"
** Welcome to Ctffind **

Version : 4.1.8
Compiled : Jun 7 2017
Mode : Scripted

Input image file name :
Error: Blank answer in scripted mode, exiting..."

So I am not sure what the problem is at this point. I haven't tried using the program directly from the command line yet. I will do that and get back with the results.

Thanks and regards,
Kavita

jerome Wed, 2018-03-28 07:53

Hello,

I've the exactly same problem

I'm under SGE gridengine cluster debian strectch
relion 2.1 up to date and Ctffind 4.1.10

I try to execute directly on a cluster node

relion_run_ctffind --i MotionCorr/job040/corrected_micrographs.star --o CtfFind/job041/ --CS 2 --HT 300 --AmpCnst 0.1 --XMAG 10000 --DStep 3.54 --Box 512 --ResMin 30 --ResMax 7.1 --dFMin 5000 --dFMax 50000 --FStep 500 --dAst 100 --use_noDW --ctffind_exe /opt/ctffind4.1.5/bin/ctffind --ctfWin -1 --is_ctffind4

But I had no result at all .. with the same message

Using CTFFIND executable in: /opt/ctffind4.1.5/bin/ctffind
to estimate CTF parameters for the following micrographs:
*
*
Estimating CTF parameters using Alexis Rohou's and Niko Grigorieff's CTFFIND4.1 ...
0/ 0 sec ............................................................~~(,_,">
WARNING: skipping, since cannot get CTF values for
WARNING: skipping, since cannot get CTF values for

when I try to replace by
relion_run_ctffind --i MotionCorr/job040/Micrographs/Falcon_2012_06_12-17_23_32_0.mrc --o CtfFind/job041/ --CS 2 --HT 300 --AmpCnst 0.1 --XMAG 10000 --DStep 3.54 --Box 512 --ResMin 30 --ResMax 7.1 --dFMin 5000 --dFMax 50000 --FStep 500 --dAst 100 --use_noDW --ctffind_exe '/opt/ctffind4/bin/ctffind' --ctfWin -1 --is_ctffind4
Using CTFFIND executable in: /opt/ctffind4/bin/ctffind
to estimate CTF parameters for the following micrographs:
* MotionCorr/job040/Micrographs/Falcon_2012_06_12-17_23_32_0.mrc
Estimating CTF parameters using Alexis Rohou's and Niko Grigorieff's CTFFIND4.1 ...
1.25/1.25 min ............................................................~~(,_,">
Segmentation fault (core dumped)

To me, it seem that name was not passed to ctffind ... when I use corrected_micrographs.star ...

I'm lost :)

Best regards

jerome Wed, 2018-03-28 08:50

Hello,
I found one prob in my CLI if I erase --UseNoDW my job in line finish

relion_run_ctffind --i MotionCorr/job040/Micrographs/*.mrc --o CtfFind/job041/ --CS 2 --HT 300 --AmpCnst 0.1 --XMAG 10000 --DStep 3.54 --Box 512 --ResMin 30 --ResMax 7.1 --dFMin 5000 --dFMax 50000 --FStep 500 --dAst 100 --ctffind_exe '/opt/ctffind4/bin/ctffind' --ctfWin -1 --is_ctffind4
Using CTFFIND executable in: /opt/ctffind4/bin/ctffind
to estimate CTF parameters for the following micrographs:
* MotionCorr/job040/Micrographs/Falcon_2012_06_12-17_23_32_0.mrc
Estimating CTF parameters using Alexis Rohou's and Niko Grigorieff's CTFFIND4.1 ...
1.30/1.30 min ............................................................~~(,_,">
Done! Written out: CtfFind/job041/micrographs_ctf.star

And after that I try with corrected_micrographs.star without the --UseNoDW and It ( seems to ) work(s)

relion_run_ctffind --i MotionCorr/job040/corrected_micrographs.star --o CtfFind/job041/ --CS 2 --HT 300 --AmpCnst 0.1 --XMAG 10000 --DStep 3.54 --Box 512 --ResMin 30 --ResMax 7.1 --dFMin 5000 --dFMax 50000 --FStep 500 --dAst 100 --ctffind_exe '/opt/ctffind4/bin/ctffind' --ctfWin -1 --is_ctffind4
Using CTFFIND executable in: /opt/ctffind4/bin/ctffind

Now I will try with relion 2.2

Best Regards

jerome Wed, 2018-03-28 08:57

Hello,
Sorry I would say Relion 2.1 ... but I works when the option "use micrograph without dose-weighting" is disabled

I'm only an I.T who works in genetics laboratory ... I'm not sure that the results are good ... but .. for me it works.

I'll be back later after contacting researcher who use relion and ctffind :)

Best regards