| |
| Readme |
| Data
Retrieval |
| NMR
Draw Instructions |
| Macros |
| Macros |
| Macros |
| Macros |
The general procedure for
processing NMR data is as follows:
1. be aware of the extent of the experiment... is it
done, when will it be done, etc. you can determine this
by checking the acqu2s files in the directory where the ser
file is... more acqu2s /$TD this will search for the
TD string and give you the value for the time domain in the
indirect dimension. This is the value of yN in the conversion
macro and when the experiment is complete, will be equal
to 2 * L1.
2. retrieve your data from the spectrometer and save
it on an sgi in the lab where you have a data directory.
(see DATA_RETRIEVAL for instructions on this)
3. copy the necessary conversion and processing macros
from some you already have or a common source like this
one. give it a name that associates it with your experiment
a good practice is consecutively naming experiments
you've run. e.g. the experiment numbers on the 800 might
include 100 - 108, then if you switch to the ge600, you could
start at 109 and continue from there. Name your saved
raw data (the ser file you bring over), the header file, the
final nv file, and any conversion and processing macros used
the same way with a different suffix. e.g. 108_sl12_hmqc.raw
108_sl12_hmqc.hdr 108_sl12_hmqc.nv 108_sl12_hmqc.conv 108_sl12_hmqc.com
4. modify the conversion macro to fit your data (see
instructions in README files where the conversion macros
are located) and run it.
5. use nmrDraw to obtain zero and first order phase
values (see instructions in NMRPIPE/general_processing_instructions/
NMRDRAW_Instructions)
6. modify d1 nmrpipe processing macro (that processed
the first dimension) to fit your data and run it.
7. use nmrDraw to obtain phase values for the second
dimension
8. modify the d1d2 processing macro and run it.
9. If a 2D, convert the ft2 file to nmrview using the
following: nmrPipe -x -in file.ft2 | pipe2xyz -nv -x
-out file.nv -verb -ov
10. If a 3D or 4D, run the processing macro to process
the 3rd and 4th dimensions (you shouldn't need phase
values if you set the constants correctly when you set up
the experiment... see the README files for the specific
macros for each experiment for details.) Then convert
to nmrview: xyz2pipe -x -in file%03d.ft3|pipe2xyz -nv -x -out
file.nv -verb -ov or xyz2pipe -x -in file%03d.ft4|pipe2xyz
-nv -x -out file.nv -verb -ov
11. intermediate files (.fid, .ft2, .ft3, and .ft4 files)
are temporary and take up space, so I usually delete
them each time I'm done processing or I call them all temp
and overwrite them each time I process data. |
To retrieve data for
processing from the spectrometers: either
a) telnet to the spectrometer (telnet 800.hhmi.umbc.edu,
telnet nmr, or telnet ge600); give login and password
then go to the directory that contains your raw data (probably
something like /u/data/yourname/nmr/directoryname/expnumber)
then remote copy the ser (serial) file containing your
raw data and the header (acqus) file containing the parameter
values: rcp ser cobalt:/data/d15/dana/tocsy_sl12.raw
rcp acqus cobalt:/data/d15/dana/tocsy_sl12.hdr logout of spectrometer
or
b) go to the directory in which you want the data ftp
(file transfer protocol) to the spectrometer (ftp 800.hhmi.umbc.edu)
give your name and password cd to the directory that contains
the ser and acqus files retrieve the data get ser tocsy_sl12.raw
get acqus tocsy_sl12.hdr type bye to exit ftp
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
NOTE: you can determine the extent of the experiment
by checking the other header files: more acqu2s and
searching (type / then the string you want to search for)
for $TD will give you the number of completed total
points in the indirect dimension e.g. if you set L1 to 256
and the experiment is not quite done yet, type more
acqu2s on the spectrometer in the shell that contains
your ser file then type /$TD that will take you to the TD
parameter which when the experiment is complete will
read 512 (2*L1=1td) This can also be done for 3D and 4D experiments
(keeping in mind that since there's no parmode 4, the value
of 2td in a 4D is L2*2*L3*2.) |
To start nmrDraw, type
nmrDraw in a unix shell where your data processing will take
place. Some people use an ampersand (&) after the command
to start the program. All this does is run the program
in the background allowing you to use the unix shell in which
you started the program. click right on File --> Select
file in menu, click left once on file name, then once on read,
once on done. typing D will draw the dot plot typing h will
give a 1D horizontal slice drag the slice down to y=1
(the bottom of the screen) typing N will bring up the nmrpipe
command window type in the command line: cs -ls 72 -sw
(if the parameter DECIM in the header is not a mult.
of 16, you need a value other than 72 here.... type z to get
a zoom box. move it to the bottom left corner where
the data begins (around 60 - 75 along x) then see where the
fid first crosses zero just before it increases dramatically
for the first time (this value will probably be 60,61,62,70,71,or
72) and this is the value you will use) right click on functions
and scroll to cosine bell left click execute right click
on functions and scroll to fourier transform left click on
execute left click on done change amplitude of y axis by holding
middle mouse button and dragging up or down in the purple
peripheral region click on phase on change p0 to phase the
area of interest use the left mouse button and click
on the purple area below the area of main interest...
this will serve as your p1 pivot (p1 phase change will leave
this area unaffected) change p1 to phase the rest of
the spectrum these are the values of p0 and p1 you will use
in your macro
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
When reading in ft1 files (just the rows are processed
and you want to get phase values for the 2nd dimension),
do the same as above except type v instead of h to get a vertical
1D. Move the vertical slice to be of an area with a
good intense signal. nmrpipe process it the same way except
do not do the cs command and between the cosine bell and ft
commands, do a zero fill. Once that vertical slice is
processed increase it to the desired amplitude, then type
a to append another vertical slice. Move this vertical
slice to another area of the spectrum preferably the
opposite side and again look for an area with intense signal.
Process the same way. Use p0 and p1 to try to phase
both spectra together. These are the values of p0 and p1 you
will use for the second dim. |
|
|
|
