Project and Subtract : Compute re-projections of map to be subtracted, and subtract them from the original images.

Usage in command line

	sp_signalsubtract.py projsubtract --origparts=particle_stack --map2subtract=map_to_subtract --projparams=projection_parameters --outdir=output_subtraction_directory --nmontage=number_of_examples --inmem --saveprojs --stats --nonorm --verbosity=verbosity_level

sp_signalsubtract exists in both MPI and non-MPI versions (Running MPI version does not require –MPI flag).

sp_signalsubtract comprises the steps needed to subtract masked regions from a map in order to focus on the remaining features from that map. There are six steps for this process (five within SPHIRE):

1. Average and low-pass filter a map for segmentation

2. Locate regions to be subtracted

3. Mask a soft mask of the region to be excluded

4. Compute re-projections of map to be subtracted, and subtract them from the original images :

mpirun -np 16 sp_signalsubtract.py projsubtract --origparts=particle_stack --map2subtract=map_to_subtract --projparams=projection_parameters --outdir=output_subtraction_directory --nmontage=number_of_examples --inmem --saveprojs 

The flag “–nmontage=number_of_examples” will write examples of original/projection/subtracted images.

The flag “–inmem” will store the projection images in memory. By default, the projections will be written to disk continuously. Storing them in memory will be about 5 to 10% faster, but may require MUCH more memory. A memory estimate will be displayed in the output.

By the default, the intermediate projections will be deleted after subtraction. To save them, use the “–saveprojs” flag.

5. Compute reconstruction of density remaining after signal-subtraction :

sp_meridien.py bdb:output_subtraction_directory#all_subtracted meridien_directory --do_final=iteration_number --voldir=output_subtraction_directory

This step will compute a reoncstruction using Meridien, i.e., with “smear” information. The iteration number, for the refinement in “meridien_directory”, is specified by the flag “–do_final”.

Note: If the “–voldir” flag is not specified, the resulting reconstructions will be written to the Meridien directory, which may overwrite existing files.

Example when using the optional flag “–nmontage”, and low-pass filtered. At the left is a raw image. In the middle is a projection. At the right is the difference image. If there is a high-contrast feature in the difference image, something went wrong: probably a map with an incorrect intensity range was used.

6. Center map of the remaining density

--origparts

Original particle stack before signal subtraction. (default None)

--map2subtract

Map whose projections will be subtracted from the original images. (default None)

--projparams

Angles and shifts, from Meridien. (default None)

--outdir

Directory where outputs will be written. (default required string)

--verbosity

Controls how much information will be written to the screen. (default 2)

--nmontage

This number of original, projections, and subtracted images will be written to disk. (default 0)

--inmem

Flag to store projections in memory. (default False)

--saveprojs

Flag to save intermediate projections. (default False)

--stats

Flag to save normalization statistics. (default False)

--nonorm

Flag to skip normalization. (default False)

docnormstats.txt

Normalization statistics

all_subtracted

Combined BDB stack of signal-subtracted images

stktestmontage.hdf

Montage of test images

Bai XC, Yan C, Yang G, Lu P, Ma D, Sun L, Zhou R, Scheres SHW, Shi Y. (2015) “An atomic structure of human γ-secretase.” Nature 525: 212-217. doi: 10.1038/nature14892.

Tapu Shaikh

Based on a protocol developed by Zhong Huang and Pawel Penczek

Category 1:: APPLICATIONS

sphire/bin/sp_signalsubtract.py

Beta:: Under evaluation and testing. Please let us know if there are any bugs.