plotphase.com {tubename}
Output: {tubename}.phase
View file with gnuplot - phases should be 0 or 180.
5. Use acf to determine repeat length. Make {tubename}.acf control file.
acf {tubename}
Output {tubename}.acf.log
6. Go back to step 1. Repeat until best repeat found.
7. Make projection plot of image
prjpltfat
output: prjplt.ps
7. Determine CTF of image: Need pltctf.def file.
sctravgft
pltctfx {tubename}
8. Using mathCAD and
my own programs, determine helical symmetry, or at least determine a set
of 2 or 3 most probable symmetries.
nltablk 80 2.534
Outpute {tubename}.tbl
3. Change data format of above table to useful format:
nlt.com {tubename}
lnl.com {tubename}
Outputs nltable.{tubename}, {tubename}.lnl
4. Edit hlxs.com. Use {tubename}.lnl for list of l n l for symmtery.
hlxs.com {tubename} {peakheight limit}
Use (say) 1800 for peakheightlimit
Outputs - {tubename}.src
5. Edit srch.com for searching - use crude search at first for tilt and xshift.
srch.com {tubename}
Output: {tubename}.rdl - residual of best fit
srch.log - log file of all searches
{tubename}1.src List of all good peaks
6. Re-run srch.com , with fine limits, around best tilt and xshift. Run again using {tubename}1.src as input file:
srch.com {tubename}1
Output: {tubename}11.src
srch.log
{tubename}1.rdl
7. When happy with tilt and xhift (both "reasonable" values, good number of peaks included in output .src files), edit hlxmk.com to make .hlx file:
last line: 1 50 130 for 130 gridunits - will extract within 1st ctf.
hlxmk.com {tubename}
Output file: {tubename}.hlx
8. Edit {tubename}.hlx - copy lines from header.hlx. Edit to account for actual tilt (omega) and xshift - determined from adding up .rdl files.
9. Extract layer lines -use hlxfl
hlxfl
(will ask for control file - .hlx file. Also scale factor : 1)
Output: {tubename}.nea
{tubename}.far
10. Edit {tubename}.halfit for calculation of near/far phase residual. This file uses the same format as hlxs.com, so can use {tubename}.lnl as input.
Only enter layer lines that you know are good:
(0,0),(1,0), (-1,1),(0,1),(1,1),(0,2),(1,2),(1,3)
hlxresidual.com {tubename}
Output: hlxresidual.log - near/far phase residual.
.rdl file - not really useful.
11. Judge quality - phase residual should be better than 65 deg. If bad, try other possible helical symmetries, maybe try other repeats, etc. HOWEVER, a low-defocus image is inherently worse, so if defocus is low, be more leniant.
12. can run nfavg and convavg to average near/far sides and convert to CT format for unit cell calculation.
nfavg
convavg
Output {tubename}.avg
{tubename}ct.avg
13. Determine un it cell parameters
Output {tubename}bin.lg2
15. Edit lglst.cnt to match tubename.
lglst lglst
Output lglst.gnu
lglst.dat
16. View MRDD with gnuplot:
(in gnuplot)load 'lglst.gnu'
set term postscript
set output 'mrdd.ps'
replot
If quality good, can
stop here for awhile until several images have been collected or several
datasets processed.
Next steps: Extraction
of data to 300 grid units, application of ctf, and (if yet another unque
helical symmetry) extraction to little-g and fitting of tube to average
in order to determine radial shift.
1. Re-run nltablk with limts of 200 for Bessel order and 300 gu:
nltablk 200 2.534
2. Re-run nlt.com and lnl.com
3. Edit hlxmk.com to output file to 300 gu.
hlxmk.com {tubename}
output {tubename}300.hlx
4. Edit {tubename}300.hlx - add info from header.hlx. Run hlxfl
hlxfl
Output: {tubename}300.nea
{tubename}300.far
5. Run nfavg to make .avg file - will be used ONLY for ctfplotr.com
nfavg
Output {tubename}300.avg
6. Edit ctfplot.cnt to match predetermined ctf then run ctfplotr.com
ctfplotr.com
Output ctf{tubename}300.nea
ctf{tubename}300.far
ctfplot1.plt, ctfplot2.plt,ctfplot3.plt,ctfplot4.plt
7. Run nfadd.com to add near and far sides
nfadd.com {tubename}300 1 1
Output ctf{tubename}300.avg
8. Edit divctfbp1.com to divide by ctf2 but not zero out any data- no bkg yet!
Run with snc and rsnc equal to 0.0
divctfbp1.com
Output: call{tubename}300.lim
9. Edit hcut.com to cut above file at 130 gu resolution.
Outputs {tubename}130.lg2
11. Edit ltlg_reindex.com to reindex tube from determined symmetry to symmetry of reference dataset.
ltlg_reindex.com
Outputs: {tubename}130r.lg2
12. Edit matchltlg.com to match layer lines between ref and tube dataset.
matchltlg.com
Outputs: {tubename}130rm.lg2
avgac130m.lg2
13. Edit hlxfitlg.com to fit reindexed tube with reference. Run it:
hlxfitlg.com {tubename}130rm
Outputs: hlxfitlg.log
{tubename}130rm.fit
14. Examine above .log file to find minimum residual. Make table of this residual, along with radial shift (from 200-450), phi, z, rscal, R and scale factor.
15. Edit hlxfb.cnt file to alter radius by 2 A intervals, so (say) go from 252 to 452. Go back to step 10 until radius at minimum residiual is found. Also, try rotating tube 180 degrees and see if this helps.
Note- approximate radial
shift can be predetermined from mrdd plots.
Minimum phase residual provides another handle for determining quality of data. Radial shidt determined in this set will be used for reindexing the average (reference) dataset for tube straightening (step IV).
Note also that files
produced up to here will NOT be used for actual averaging - the data must
first be straightened. Up to now, the processing has been assessment of
quality, determination of helical symmetry and relative radial shift of
tube.
ltlg_reindex.com
Output avgac300r.lg2
2. Run matchltlg.com to match layer lines so there will be no problem for little-g to big G conversion. Also changes radius to match that of tube under study.
matchltlg.com
Outputs avgac300rm.lg2
3. Run .com to make control file for hfbr.
Outputs
4. Edit hfbr.cnt to work on avgac300rm.
hlxbfr-ascii hfbr
Output: avgac300rm.cor
5. Run convavg.com to convert above .cor file to Unwin format
convavg.com avgac300rm
Output avgac300rmun.cor
6. Edit and run hcut.com to cut avg file to 1st CTF of tube under study (say 130 g.u.)
hcut.com
Output avgac130rmun.cor
7. Make {tubename}a.box file - Unwin box file format. Follow his protocol for straightening tube using file created in step 7 as reference.