> i) How to display more than one contour level for a given map?
you make internal copies of a map:
MAIN> copy map 1 2 MAIN> image color 200 map 2 dens 1.3 box 15 15 15
this copy just copeis a header and not the data arrray. You can deal with 10 maps simultaneously each at a different contouring level.
try
MAIN> show map
or
MAIN> show map 2
> ii) Can you explain me how to use the menu on the graphic screen, if > it is in the manual I could not find it.
x-menu and it is in appendix A within DIALOG MODE ACCESSORIES where you will find also the keybord and mouse usage descriptions.
dialoge mo de?> iii) Can you give some example of using the selection menu during
First read the description in x-menu and afterwards also the SELECT chapter in command reference manual, and afterwards try to do the following:
Menu is relatively complicated since it allows you to 3 to power of 4 combinations. Try first the simplest:
choose a molecule (molecule is in this combination a network of
and the selected molecule should be marked with white stars
to learn more set the level of MAIN information higher
MAIN> set lev 3
and then enter the dialog mode and watch the syntax appearing in the MAIN window after clicking menu options. Note that the buttons are flags only - they do not do anything. Action is caused by pressing non buttoned menu keys.
energy calculations and minimizations in the dialog mode
You need to define 2 keys 'active' and 'passive' (KEY\_PASI),
and all atoms included in active will be copied to the key passive in addition each atom of a covalently attached residue to key active will be included in passive as well. Thereby the key active is anchored to the rest of a molecule.
If you have already DEFINED your energy terms, press SHOW\_ENE to see the energy terms of the two selections. Key active is interaction with key passive. Active atoms are allowed to move while passive can move only if they are included in the key active as well.
If energy terms are not defined go to page 4 and click DEFINE\_A. There may be some warnings comming because some atoms are missing. Take serious the one related to RESIDUE or ATOM not found and the ones related to bond and angle terms.
See what you can do with a RENAME command to adjust your system. very common case:
MAIN> rename atom CD select residu name ILE .and. atom name CD1 end
MINIMIZE the selections by clicking MINIMIZE (by default it does 50 steps)
The dials window cannot be read in stereo.
> I found some options in the examples which are not in the > documentation. For instance creating continuos surfaces to > represent maps. Is it possible to create trasparent surfaces?
only as maps - not as connoly surfaces
but be aware of the hardware limitations:
MAIN> image map 2 dens 1.0 box 10 10 10 area
is a code to generate a map surface (or the MAP\_AREA menu item).
clicking the REFLECT mode in the image window turns the reflectance mode on then click the Amap\_2 for example and rotate dials - transparency is the dial number 4 (the same as scaling).
there is still some problem with it - in some directions transparency still has some problems.
What algorithm are you using for energy minimization?
Konjugated gradient.
From Ren Bin in Sweden - averaging
> what's the purpose of RHOGEN? what is the difference > between the main map and the protein map?
The purpose of this file is replace the FFTFC program that was used before to calculate structure factors from atomic model. RHOGEN.COM generates electron density from atomic model through the whole unit cell and writes the map in the Lyn Ten Eyck format readable by PROTEIN.
> In RHOGEN, I need a file called "set\_cell\_box". what > kind of file is it? How to determine the input values in > this file?
RHOGEN requires layout (0 0 0 to the end of the cell)
> In RHOGRN, the origin of the box was set to "0 0 0" > Is this a general situation? How to determine it?
Yes it should always be 0 0 0.
> (4). In the command file you sent to me, the symmetry > operations repeated 3 times, using matr 2, matr 3 and matr > 4. It seems that the number of repeat corresponds to the > number of the crystal symmetry operations. Is this correct? > In my case, I have space group I4(1). Should I repeat 7 > times, from "matr 2" to "matr 8"?
Yes the number of reeats cooresponds to the number of symmetry operations that are carried out one by one. The case I have sent you has C2 group.
The current version generates electron density from atomic model appyling all crystal symmetry operation:
MAIN> make map 1 atom dens select all end symmetry
> Another problem is how to determine "tran 0 0 0". Is > this the translation of the crystal symmetry operation? If > I have a symmetry operation like: "-x+1/2, -y+1/2, z+1/2", > should I specify it as "tran 0.5 0.5 0.5"? Why did you use > "0 0 0" in all of the symmetry operations?
The 'tran 0 0 0' is an additional translation applied along the crystal axis. It is a MAIN syntax requirenment. It simply means that no additional translation will be applied. Translational components can only be integer numbers and never 0.5. The 0.5 or similar shifts should be included already in the symmetry operators.
> (5). After RHOGEN, I will calculate the density map > *.diy. In the calculation, I need a file called "br3.pro"? > Is this a fobs file (Fc) calculated from PROTEIN using FFT? > If not, how to generate it?
br3.pro is a PROTEIN reflection file.
> (6). In your last message, you suggested that it's better > to use RMS.COM to generate the rotation matrix that will be > used in averaging. How to incorporate the > "set\_tran\_*\_to\_*.com" into the command files of the new > version? should I do averaging in two steps (like the old > version): ROTA\_MAPS.COM and AVER\_MAPS.COM ?
You can do it as you wish. The previous version stores intermediate results to disk and does averaging within an additional step. The example I have sent you no intermediate results (rotated maps) are beeing stored.
The set\_tran*to*.com defines parameters for the
> make map nn from mm rotate ...
so that you can still use the 'old' syntax for the map rotations.
On the fly calculations of parameters are neccessary if your model is being refined during averaging, because model transformations slightly effect
the rotational and translational parameters.
> (7). In GEN\_CELL.COM, file "head.rmap" will be used. I have > edited a head.rmap, but I don't know whether it is correct? > There are different kinds of head.rmap in the manual. I > don't know which one I should follow. Could you send a > example to me?
You should use your own head.rmap. The head.rmap contains the symmetry parameters, cell constants and layout of the map. It is always case specific - it means that I can not create a header for you.
> Sorry to ask so many questions. I wish you wouldn't feel bored.
Thanks for good wishes.
Rick, surface point file and end of selection
> Writing out surface file. Reading back there are holes > in it. If you display immediately after calculation it is > ok, so apparently not all of the surface is written to the > file.
The default should be all points (but I am never sure). All points mean contact + reentrant. I guess that some default must be wrong. Try:
MAIN> image col 160 poin surface all
and compare it with
MAIN> image col 200 poin surface
or
MAIN> write poin surface all
and
MAIN> write poin surface
> If you have a segment name 'E', you cannot make the > command: select seg name A .or seg name B .or seg name C > .or seg name E end because E is confused with end (or > something like this).
END can be abreviated to an E. (First the interpreter searches for end, then logical operations (.and. .or. .not. and brackets) and afterwards starts to interpret tags. You can get around with this:
> ... select segm name A B C D E* end
which includes atoms of segments A, B, C, D and all segments starting with E.
To be sure that you have set the procedure properly check each step. Inspect all the first time averaged maps (not only the final result and its r-value).
Look at messages coming out of MAIN and try to understand them. Comparing numbers of converted, rotated and set grid points tell you a lot about consistency of your masks, and the procedure in general.
The numbers coming out of MAKE_MAP_CELL> tell you how many of the density points of each mask have been brought to the P1 environment.
> write over file xx refl fobs complex defined
should be OK. PHASE is not a command word there (POLAR is what writes amplitudes (fcalc) and the phases.) Complex is another possibility. Try :
> write refl ?
to get the ideas about the possible command words.
There is no need to write everything in an X-plor map format.
First: the main number format is under VMS much smaller
> write over file average.rmap map 1 number
if you are happy with 8 bit precesion you can store a map with:
> write .... map 1 char
or with ALPHA instead of the CHAR format . CHAR maps are not transferrable to other machines, with the alpha form is computer system independent (but is only MAIN readable ).
second: use MAKE syntax to create a smaller map for visual inspection:
> make map 2 from 1 around 8 select segment name ... end init 9999 > make map 2 from 1 copy > write file xx.rmap map 2 xplor/alpha ....
the around 8 means create a box 8 grid points outside your selection. If the averaging procedure has converged , then it should be in principle enough to work with a single model only.
> still I dont know how to avoid this non unique selection E > > rename segm MOL4 sele segm name D end > MAIN> > rename segm MOL5 sele segm name E end > FINDCOM> COMMAND IS NOT UNIQUE **E en** > MAIN> > rename segm MOL6 sele segm name F end > MAIN>
The segment name E is a pain of the select sentence. MAIN thinks that it is an abbreviation for END. See in the manual user questions or users corner (at the very end) there should be something about this. A quick suggestion: Use
> rename segm MOL5 sele segm name E* end
> in the step rms_fit.com there are some things in the > log file that I don't understand. What is this SPECIAL STEP > and is the message "CAN NOT CONTINUE" a normal termination > of the process? > The special step gives no improvement in about 50% of the cases > and improvement in the rest as you can see...
This was OK. The message only means that the procedure did not numerically converge, however the results could not be improved either. What you should care about is the last value in the line
> RMS_FIT> PARAMETERS -7.145 -66.704 -99.066 -84.176 97.608 48.024 0.3329
^^^^^^
This number is the RMS deviation. If it is unreasonable high (above 5 or so) then you should offer the fit procedure some initial values to improve convergence. As I see everything is OK with your case.
> rms coor all sele ... end sele ... end init 0. 180. 0. 70. 8. 90.
or whatever (x, y' and x'' rotations plus 3 translations).