Oct. 4, 1999
This protocol is used to average several tubes of several different
helical symmetries. It involves
1) reindexing all tubes to a common helical symmetry
2) determining backround (ice) FFTs and reindexing them to a common
symmetry
3) adding ctf values to reindexed files by "reindexing" ctf
4) adding all Big G files by vector addition, so that signal is allowed
to accumulate over "noise"
5) Near and far sides kept separate until very end so that near/far
stats can be determined and quality of data assessed.
Assumed- already have fitted all near/far averaged little-g files to a common symmetry and reference. (ie have made tubename.hlxshiftlg control files)
Assumed- already have made ice file stuff
Assumed - already have optimised CTF for each tube
Commands you enter are in red courier
1 Run ssadd.com on ctfb{tubename}ac300.nea and .far files:
ssadd.com {tubename}ac300.nea
ssadd.com
{tubename}ac300.far
Makes ctfb{tubename}ac300add.nea and .far files.
These files have amplitudes multiplied by |CTF|, and backgrounds squared and multiplied by |CTF|. The ctf column is replaced by (ctf)^2.
The background will not be used. Could therefore also run this on ctf{tubename}ac300 files.
The
CTF will also not be used.
2
Run farfliplim.com to "flip" far side. (ie add 180 deg to phase if n is
odd)
farflip.com
ctf{tubename}ac300add
Outputs
file ctf{tubename}ac300flip.far file.
3 Edit hlxfb.cnt file to work on files created in steps 2 and 3.
Output files ctf{tubename}ac300nea.lg2
ctf{tubename}ac300farflip.lg2
hlxfb-ascii
hlxfb
4 Reindex .lg2 files to common symmetry, using already-created littleg_reindex.com
Output files ctf{tubename}ac300near.lg2
ctf{tubename}ac300farflipr.lg2
ltlg_reindex.com ctf{tubename}ac300nea
ltlg_reindex.com
ctf{tubename}ac300farflip
5 Edit {tubename}.hlxshftlg to work on files created in step 5.
Save as {tubename}ac300near.hlxshftlg and {tubename}ac300farflipr.hlxshftlg.
Output files ctf{tubename}ac300nearshft.lg2
ctf{tubename}ac300farfliprshft.lg2
hlxshftlg.com {tubename}ac300near
hlxshftlg.com
{tubename}ac300farflipr
6 Edit and run matchltlg.com to match reindexed little g files with reference file.
Run on both nea and far files.
matchltlg.com
Outputs ctf{tubename}ac300nearshftm.lg2
ctf{tubename}ac300farfliprshftm.lg2
Purpose: 1) Ensures that all layer lines will work in reverse Fourier-Bessel transform
2) Makes radius of tube match that of reference tube.
7 Edit hfbr.cnt to work on reindexed, shifted, matched little-g near and far files.
hlxfbr-ascii hfbr
Output files: ctf{tubename}ac300nea.cor
ctf{tubename}ac300far.cor
8 Run convavg.com to convert above files to Unwin Big G format.
convavg.com ctf{tubename}ac300nea
convavg.com ctf{tubename}ac300far
Output files: ctf{tubename}ac300neaun.cor
ctf{tubename}ac300farun.cor
9
Edit
above .cor files with emacs: Use "replace" command to replace AVG with
NEA and FAR as appropriate.
10 Edit ctfplotreindex.cnt file - use same parameters as from ctfplot.cnt for particular tube. Add proper helical symmetry info.
ctfplotreindex.com
Output files: ctf{tubename}ac300neaunc.cor
ctf{tubename}ac300farunc.cor
Purpose
-adds ctf column to files, should be proper ctf according to original helical
symmetry.
11 Now start work on ice files
Edit {tubename}.hlx file to extract ice .fft rather than tube .fft file.
hlxfl
Output: {tubename}ice.nea
{tubename}ice.far
12 Edit ctfplot.cnt file to work with ice nea and far files.
ctfplotr.com
Generates
ctf{tubename}ice.nea
and .far files
13 Run nfadd to add nea and far sides and multiply by ctf.
\public\bin\rice\nfadd
ctf{tubename}ice.nea
ctf{tubename}ice.far
ctf{tubename}ice.avg
1 1 0.0
Generates
ctf{tubename}ice.avg
14 Edit hlxfb.cnt to work with above lim file, and make little g file.
hlxfb-ascii hlxfb
Generates
ctf{tubename}ice.lg2
15 Run ltlg_reindex.com to reindex above little-g file.
ltlg_reindex.com ctf{tubename}ice.lg2
Generates
ctf{tubename}icer.lg2
16 Create {tubename}ice.hlxshftlg to shift little-g file. [OPTIONAL]
hlxshftlg.com {tubename}ice
Generates
ctf{tubename}iceshft.lg2
17 Edit matchltlg.com to match ice and reference layer lines
matchltlg.com
Generates
ctf{tubename}iceshftm.lg2
18 Edit hfbr.cnt to work with matched little g file
hlxfbr-ascii hfbr
Generates
ctf{tubename}ice.cor
19 Run convavg.com to convert to Unwin's format Big G file
convavg.com ctf{tubename}ice
Generates
ctf{tubename}iceun.cor
20Edit bkgadd.cnt to add background column. Start with scale factor 1.
bkgadd.com
Check log file and adjust scale factor accordingly. Re-run.
Generates ctfb{tubename}neaunc.cor
ctfb{tubename}farunc.cor
21 Run ctfsqr - replaces ctf and bkg with ctf2 and bkg2
ctfsqr
Generates cctfb{tubename}neaunc.cor
cctfb{tubename}farunc.cor
22 Edit tubeadd.cnt file for number of tubes and each tube's scale factor.
tubeadd.com
Generates ctfbavgfar.cor
ctfbavgnea.cor
23 Edit divtcfbp1.com to divide amplitudes by ctf2. Also eliminates points below snc cutoff.
divctfbp1.com
Generates ctfbavgfar.lim
ctfbavgnea.lim
24 Edit ctfavg.nfstat to determine near/far phase residuals for various resolution shells.
nfstat.com ctfavg
Generatesnfstat.log
file. Examine log file for results.
25 Edit twofold.cnt file. Run hlx2fld to determine % of data included in various resolution shells.
hlx2fld
twofold
26
Repeat steps 23-25 until desired snc cutoff determined. Note % data
included in shells - this is important for determining snc cutoff of near/far
averaged file
27 Run nfadd to average near and far averaged files.
/public/bin/rice/nfadd
Generates
ctfbavgall.cor
28 Run divctfbp1.com on ctfbavgall.cor. Run hlx2fld. Repeat until desired % data inclusion is reached.
Generates
ctfbavgall.lim
29 Finally, edit htrunc.com to truncate data beyond set resolution.
htrunc.com
Generates ctfbavgall12A.lim
Use this file to make little g file and eventual map.