Skip to content

Mlf_align2d

Jump to bottom
Adrian Quintana edited this page Dec 11, 2017 · 1 revision

mlf_align2d

Purpose

This utility allows you to perform (multi-reference) 2D-alignment, using a maximum-likelihood target function in Fourier Space (MLF). The mathematical details behind this approach are explained in detail in


Scheres et. al. (2007) Structure, 15, 1167-1177

Please cite this paper if this program is of use to you!

This program is run in a very similar way as the ml_align2d program. Please refer to the MLalign2D help page for a listing of all their common options and a description of the output files. A parallel version of this program has been implemented in mpi_mlf_align2d, which is similar to the Mpi_MLalign2D program.

Important Notes

Before using this program, please read the very important NoteOnCtfGroups and NoteOnMlfConvergenceBehaviour.

Usage


$ mlf_align2d ...

ml_align2d parameters (follow MLalign2D for a listing of all ml_align2d parameters)

But note that:

  • `` is used automatically.
  • `` does not work for MLF as the initial noise is estimated automatically from the data.

MLF-specific parameters

  • `` A two-column ASCII file with the image names as in the input selfile and the corresponding CTF-parameter filenames of each image. Note that the input selfile may be a subset of the images in the CTFdat file, but that the CTFdat file should contain all images in the input selfile, and their names should be exactly the same. Also read the importantNoteOnCtfGroups
  • OR:`` Instead of giving the CTF parameters via the-ctfdat option, one may choose to ignore CTFs by giving this option.
  • -search_shift [double3] = The integrations over the origin offsets are limited to + or - this value in the X and Y directions.
  • `` Use this flag to multiply all estimated spectral SNRs by a factor (between 0 and 1!). This may be useful in cases with few or very noisy images, in order to prevent over-estimation if the SNR at higher frequencies.
  • `` By default, the images are assumed to be phase-flipped. Use this flag for unflipped images.
  • `` If providing references using the -ref option, they are assumed to be CTF-corrected. Use this flag if they are not CTF-corrected.
  • -low [Ang999] = Exclude the lowest resolution Fourier pixels from the probability calculations.
  • -high [Ang0] = Exclude the highest resolution Fourier pixels from the probability calculations in any iteration.
  • -ini_high [Ang0] = Exclude the highest resolution Fourier pixels from the probability calculations in the first iteration.
  • -pixel_size [Ang1] = Only for the-no_ctf option, give the pixel size in Angstroms.

Examples and Notes

A typical use of this program would be:


$ mlf_align2d -i g0u.sel -ctfdat g0u.ctfdat -nref 3 -o MLFg0u -mirror -reduce_snr 0.5

Which would perform a 2D MLF-refinement using 3 reference images, with the experimental images ing0u.sel. The program will output reference images calledMLFg0u_it?????_ref0000[1-3].xmp and a logfile calledMLg0u_it?????.log after each iteration. The amplitudes of the resolution-dependent noise model will be in ASCII filesMLFg0u_it?????_ctf?????.noise and information about the estimated spectral SNRs and the corresponding Wiener filters will be in ASCII filesMLFg0u_it?????_ctf?????.ssnr. If option-ctfdat was given, all output reference images will be CTF-corrected. The spectral SNRs are calculated using CTF-affected images, which are stored asMLFg0u_cref?????.xmp and these are overwritten each iteration.

The format of the CTFdat file is for eaxmple:


images/img00001.xmp ctfs/ctf_group_1.ctfparam
images/img00002.xmp ctfs/ctf_group_1.ctfparam
images/img00003.xmp ctfs/ctf_group_1.ctfparam
images/img00004.xmp ctfs/ctf_group_2.ctfparam
images/img00005.xmp ctfs/ctf_group_2.ctfparam
images/img00006.xmp ctfs/ctf_group_2.ctfparam
images/img00007.xmp ctfs/ctf_group_3.ctfparam
images/img00008.xmp ctfs/ctf_group_3.ctfparam
images/img00009.xmp ctfs/ctf_group_3.ctfparam

for a selfile like:


images/img00001.xmp 1
images/img00002.xmp 1
images/img00003.xmp 1
images/img00004.xmp 1
images/img00005.xmp 1
images/img00006.xmp 1
images/img00007.xmp 1
images/img00008.xmp 1
images/img00009.xmp 1

--Main.SjorsScheres - 10 Sep 2007

Clone this wiki locally