Loading ...
Sorry, an error occurred while loading the content.

4D Linear Prediction Script

Expand Messages
  • Geoff Armstrong
    Hello All, I am having problems applying Linear Prediction to all indirect dimensions of my 4D data set. I can apply it easily to the individual projections
    Message 1 of 6 , Jun 10, 2004
    • 0 Attachment
      Hello All,

      I am having problems applying Linear Prediction to all indirect
      dimensions of my 4D data
      set. I can apply it easily to the individual projections that I
      extract from the FIDs, but once
      I apply it during the 4D processing it never seems to work. Looking
      at the result it almost
      seems like the forward Fourier transform is not applied after the LP
      step (i.e. it still looks
      like an FID in the indirect dimension). I can't seem to figure out
      what is going on, can
      anyone help? Thanks in advance! I'm including the script below for
      reference.

      Geoff

      #!/bin/csh

      #
      # 4D Processing with Linear Prediction in Y, Z and A Dimensions:
      # Process the Directly-detected X-Axis.
      # Process the Indirectly-detected Z-Axis.
      # Process the Indirectly-detected A-Axis.
      # Predict and Process the Indirectly-detected Y-Axis.
      # Inverse Process, Predict, and Re-Process the Z-Axis.
      # Inverse Process, Predict, and Re-Process the A-Axis.

      xyz2pipe -in data/test%02d%03d.fid -x -verb \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 -137 -p1 0.0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -x -ov

      xyz2pipe -in lp/test%02d%03d.dat -z -verb \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 0.0 -p1 0.0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -z -ov -inPlace

      xyz2pipe -in lp/test%02d%03d.dat -a -verb \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 0.0 -p1 0.0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -a -ov -inPlace

      xyz2pipe -in lp/test%02d%03d.dat -y -verb \
      | nmrPipe -fn LP -ps0-0 \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 0 -p1 0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -y -ov -inPlace

      xyz2pipe -in lp/test%02d%03d.dat -z -verb \
      | nmrPipe -fn HT \
      | nmrPipe -fn PS -p0 0.0 -p1 0.0 -inv \
      | nmrPipe -fn FT -inv \
      | nmrPipe -fn ZF -inv \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -inv \
      | nmrPipe -fn LP -ps0-0 \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -z -ov -inPlace

      xyz2pipe -in lp/test%02d%03d.dat -a -verb \
      | nmrPipe -fn HT \
      | nmrPipe -fn PS -p0 0.0 -p1 0.0 -inv \
      | nmrPipe -fn FT -inv \
      | nmrPipe -fn ZF -inv \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -inv \
      | nmrPipe -fn LP -ps0-0 \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 0 -di \
      | pipe2xyz -out lp/test%02d%03d.dat -y -ov
    • Ryan McKay
      Try it first without the back predictions. e.g. x lp y lp z lp a Then add back in the back predictions one at a time and see where/if it fails. However if it
      Message 2 of 6 , Jun 10, 2004
      • 0 Attachment
        Try it first without the back predictions.

        e.g.

        x
        lp y
        lp z
        lp a

        Then add back in the back predictions one at a time and see where/if it
        fails.

        However if it looks fine without back prediction .....then I would
        guess you don't need it.

        Ryan





        On Jun 10, 2004, at 12:39 PM, Geoff Armstrong wrote:

        Hello All,

        I am having problems applying Linear Prediction to all indirect
        dimensions of my 4D data
        set.  I can apply it easily to the individual projections that I
        extract from the FIDs, but once
        I apply it during the 4D processing it never seems to work.  Looking
        at the result it almost
        seems like the forward Fourier transform is not applied after the LP
        step (i.e. it still looks
        like an FID in the indirect dimension).  I can't seem to figure out
        what is going on, can
        anyone help?  Thanks in advance!  I'm including the script below for
        reference.

        Geoff

        #!/bin/csh

        #
        # 4D Processing with Linear Prediction in Y, Z and A Dimensions:
        #   Process the Directly-detected X-Axis.
        #   Process the Indirectly-detected Z-Axis.
        #   Process the Indirectly-detected A-Axis.
        #   Predict and Process the Indirectly-detected Y-Axis.
        #   Inverse Process, Predict, and Re-Process the Z-Axis.
        #   Inverse Process, Predict, and Re-Process the A-Axis.

        xyz2pipe -in data/test%02d%03d.fid -x -verb         \
        | nmrPipe  -fn SP -off 0.5 -end 0.98 -pow 2         \
        | nmrPipe  -fn ZF -auto                             \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS -p0 -137 -p1 0.0 -di              \
        | pipe2xyz -out lp/test%02d%03d.dat -x -ov

        xyz2pipe -in lp/test%02d%03d.dat -z -verb           \
        | nmrPipe  -fn SP -off 0.5 -end 0.98 -pow 2         \
        | nmrPipe  -fn ZF -auto                             \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS -p0 0.0 -p1 0.0 -di               \
        | pipe2xyz -out lp/test%02d%03d.dat -z -ov -inPlace

        xyz2pipe -in lp/test%02d%03d.dat -a -verb           \
        | nmrPipe  -fn SP -off 0.5 -end 0.98 -pow 2         \
        | nmrPipe  -fn ZF -auto                             \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS -p0 0.0 -p1 0.0 -di               \
        | pipe2xyz -out lp/test%02d%03d.dat -a -ov -inPlace

        xyz2pipe -in lp/test%02d%03d.dat -y -verb           \
        | nmrPipe  -fn LP -ps0-0                            \
        | nmrPipe  -fn SP -off 0.5 -end 0.98 -pow 2         \
        | nmrPipe  -fn ZF -auto                             \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS -p0 0 -p1 0 -di                   \
        | pipe2xyz -out lp/test%02d%03d.dat -y -ov -inPlace

        xyz2pipe -in lp/test%02d%03d.dat -z -verb           \
        | nmrPipe  -fn HT                                   \
        | nmrPipe  -fn PS  -p0 0.0 -p1 0.0            -inv  \
        | nmrPipe  -fn FT                             -inv  \
        | nmrPipe  -fn ZF                             -inv  \
        | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2  -inv  \
        | nmrPipe  -fn LP  -ps0-0                           \
        | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2        \
        | nmrPipe  -fn ZF                                   \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS  -p0 0 -di                        \
        | pipe2xyz -out lp/test%02d%03d.dat -z -ov -inPlace

        xyz2pipe -in lp/test%02d%03d.dat -a -verb           \
        | nmrPipe  -fn HT                                   \
        | nmrPipe  -fn PS  -p0 0.0 -p1 0.0            -inv  \
        | nmrPipe  -fn FT                             -inv  \
        | nmrPipe  -fn ZF                             -inv  \
        | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2  -inv  \
        | nmrPipe  -fn LP  -ps0-0                           \
        | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2        \
        | nmrPipe  -fn ZF  -auto                            \
        | nmrPipe  -fn FT                                   \
        | nmrPipe  -fn PS  -p0 0 -di                        \
        | pipe2xyz -out lp/test%02d%03d.dat -y -ov





        Yahoo! Groups Sponsor

        ADVERTISEMENT
        <yhoo0504_testa_300250a052604.gif>
        <l.gif>

        Yahoo! Groups Links

        • To visit your group on the web, go to:
        http://groups.yahoo.com/group/nmrpipe/
         
        • To unsubscribe from this group, send an email to:
        nmrpipe-unsubscribe@yahoogroups.com
         
        • Your use of Yahoo! Groups is subject to the Yahoo! Terms of
        Service.





        ------------------------------------------------------------------------
        ------------
        Ryan T. McKay, Ph.D. - Scientific Director
        NANUC - Canada's National High Field NMR Centre
        Room 105 Email: ryan@...
        NANUC Bld. Phone: (780) 492-2700
        University of Alberta Fax: (780) 492-9174
        Edmonton, Alberta Cell: (780) 920-8871
        Canada T6G 2E1 AIM: ryanmckaynmr
        http://www.nanuc.ca
      • David A. Horita
        Geoff, You may also have a problem with the last step being -in -a -out -y without the inPlace because this arrangement might over-write itself. In addition
        Message 3 of 6 , Jun 10, 2004
        • 0 Attachment
          Message
          Geoff,
          You may also have a problem with the last step being -in -a -out -y without the inPlace because this arrangement might over-write itself.  In addition to Ryan's suggestion, use different output names from the input files so you can see where the processing fails.
           
          Regards,
          David

          xyz2pipe -in lp/test%02d%03d.dat -a -verb           \
          | nmrPipe  -fn HT                                   \
          | nmrPipe  -fn PS  -p0 0.0 -p1 0.0            -inv  \
          | nmrPipe  -fn FT                             -inv  \
          | nmrPipe  -fn ZF                             -inv  \
          | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2  -inv  \
          | nmrPipe  -fn LP  -ps0-0                           \
          | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2        \
          | nmrPipe  -fn ZF  -auto                            \
          | nmrPipe  -fn FT                                   \
          | nmrPipe  -fn PS  -p0 0 -di                        \
          | pipe2xyz -out lp/test%02d%03d.dat -y -ov

            
           
          -----------------------------
          David A. Horita, Ph.D.
          Department of Biochemistry
          Wake Forest University School of Medicine
          Winston-Salem, NC 27157-1016
          Tel: 336 713-4194
          Fax: 336 716-7671
          email: 
          dhorita@...
          web:  http://www.wfubmc.edu/biochem/faculty/Horita/
           
        • Geoff Armstrong
          Thanks for the responses, the suggestions have been helpful. But I have not yet been able to get satisfactory results. Just doing the LP of the
          Message 4 of 6 , Jun 14, 2004
          • 0 Attachment
            Thanks for the responses, the suggestions have been helpful. But I
            have not yet been able
            to get satisfactory results. Just doing the LP of the untransformed
            data as per Ryan's
            suggestion worked but the results were very poor. Analyzing the
            individual files produced
            by each step of the script suggests that the problem occurs during
            the first inverse
            processing step as the result is incorrect if I inverse process and
            then reprocess (without
            any LP) any of the indirect dimensions. Could it be that the Hibert
            Transform is not
            working for my data? I've been through the options for HT, but it
            doesn't seem like any of
            them apply. My data is constant time, but does not have a 1/2 dwell
            delay. Any ideas?
            Thanks in advance!

            Geoff

            --- In nmrpipe@yahoogroups.com, "David A. Horita" <dhorita@w...>
            wrote:
            > Geoff,
            > You may also have a problem with the last step being -in -a -out -y
            without the inPlace
            because this arrangement might over-write itself. In addition to
            Ryan's suggestion, use
            different output names from the input files so you can see where the
            processing fails.
            >
            > Regards,
            > David
            >
            >
            >
            > xyz2pipe -in lp/test%02d%03d.dat -a -verb \
            > | nmrPipe -fn HT \
            > | nmrPipe -fn PS -p0 0.0 -p1 0.0 -inv \
            > | nmrPipe -fn FT -inv \
            > | nmrPipe -fn ZF -inv \
            > | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -inv \
            > | nmrPipe -fn LP -ps0-0 \
            > | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 \
            > | nmrPipe -fn ZF -auto \
            > | nmrPipe -fn FT \
            > | nmrPipe -fn PS -p0 0 -di \
            > | pipe2xyz -out lp/test%02d%03d.dat -y -ov
            >
            >
            >
            >
            > -----------------------------
            > David A. Horita, Ph.D.
            > Department of Biochemistry
            > Wake Forest University School of Medicine
            > Winston-Salem, NC 27157-1016
            > Tel: 336 713-4194
            > Fax: 336 716-7671
            > email: dhorita@w... <mailto:dhorita@w...>
            > web: http://www.wfubmc.edu/biochem/faculty/Horita/
            <http://www.wfubmc.edu/
            biochem/faculty/Horita/>
          • Ryan McKay
            If you skip the first HT does the second HT then cause poor results (i.e. is it always the first HT no matter where it is performed that is causing errors) or
            Message 5 of 6 , Jun 14, 2004
            • 0 Attachment
              If you skip the first HT does the second HT then cause poor results
              (i.e. is it always the first HT no matter where it is performed that is
              causing errors) or is it one particular dimension?

              Have you checked the raw FIDs along that dimension for a bad point?
              For example in one 3D-13C-NOESY I like to use the first two points of
              the 13C dimension can be distorted if the sweep widths are too wide.
              Back predicting these two points or using a 1/2 dwell delay prior to
              acquisition removes the fault and results in the proper spectra.

              When you say that it results in poor spectra are you referring to
              signal to noise, peak shape distortions, water suppression etc.?

              Thanks,
              Ryan




              On Jun 14, 2004, at 11:40 AM, Geoff Armstrong wrote:

              Thanks for the responses, the suggestions have been helpful.  But I
              have not yet been able
              to get satisfactory results.   Just doing the LP of the untransformed
              data as per Ryan's
              suggestion worked but the results were very poor.  Analyzing the
              individual files produced
              by each step of the script suggests that the problem occurs during
              the first inverse
              processing step as the result is incorrect if I inverse process and
              then reprocess (without
              any LP) any of the indirect dimensions.  Could it be that the Hibert
              Transform is not
              working for my data?  I've been through the options for HT, but it
              doesn't seem like any of
              them apply.  My data is constant time, but does not have a 1/2 dwell
              delay.  Any ideas? 
              Thanks in advance!

              Geoff

              --- In nmrpipe@yahoogroups.com, "David A.  Horita" <dhorita@w...>
              wrote:
              > Geoff,
              > You may also have a problem with the last step being -in -a -out -y
              without the inPlace
              because this arrangement might over-write itself.  In addition to
              Ryan's suggestion, use
              different output names from the input files so you can see where the
              processing fails.

              > Regards,
              > David
              >
              >      
              >
              >       xyz2pipe -in lp/test%02d%03d.dat -a -verb           \
              >       | nmrPipe  -fn HT                                   \
              >       | nmrPipe  -fn PS  -p0 0.0 -p1 0.0            -inv  \
              >       | nmrPipe  -fn FT                             -inv  \
              >       | nmrPipe  -fn ZF                             -inv  \
              >       | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2  -inv  \
              >       | nmrPipe  -fn LP  -ps0-0                           \
              >       | nmrPipe  -fn SP  -off 0.5 -end 0.98 -pow 2        \
              >       | nmrPipe  -fn ZF  -auto                            \
              >       | nmrPipe  -fn FT                                   \
              >       | nmrPipe  -fn PS  -p0 0 -di                        \
              >       | pipe2xyz -out lp/test%02d%03d.dat -y -ov
              >      
              >        
              >      
              >
              > -----------------------------
              > David A. Horita, Ph.D.
              > Department of Biochemistry
              > Wake Forest University School of Medicine
              > Winston-Salem, NC 27157-1016
              > Tel: 336 713-4194
              > Fax: 336 716-7671
              > email:  dhorita@w... <mailto:dhorita@w...>
              > web:  http://www.wfubmc.edu/biochem/faculty/Horita/
              <http://www.wfubmc.edu/
              biochem/faculty/Horita/>





              Yahoo! Groups Sponsor

              ADVERTISEMENT
              <yhoo0504_testa_300250a052604.gif>
              <l.gif>

              Yahoo! Groups Links

              • To visit your group on the web, go to:
              http://groups.yahoo.com/group/nmrpipe/
               
              • To unsubscribe from this group, send an email to:
              nmrpipe-unsubscribe@yahoogroups.com
               
              • Your use of Yahoo! Groups is subject to the Yahoo! Terms of
              Service.





              ------------------------------------------------------------------------
              ------------
              Ryan T. McKay, Ph.D. - Scientific Director
              NANUC - Canada's National High Field NMR Centre
              Room 105 Email: ryan@...
              NANUC Bld. Phone: (780) 492-2700
              University of Alberta Fax: (780) 492-9174
              Edmonton, Alberta Cell: (780) 920-8871
              Canada T6G 2E1 AIM: ryanmckaynmr
              http://www.nanuc.ca
            • Geoff Armstrong
              Well just an update, It seems that the first HT and LP actually work regardless of which dimension you perform it in, but they introduce so much noise into the
              Message 6 of 6 , Jun 15, 2004
              • 0 Attachment
                Well just an update,

                It seems that the first HT and LP actually work regardless of which
                dimension you perform
                it in, but they introduce so much noise into the spectrum that the
                next LP just can't cope
                with the data. So basically the quality of the spectrum degrades
                subtantially with each
                application of the LP even though the resolution is improved
                somewhat. This may just be
                a limitation of the LP with the short data sets I am using. Thanks
                for all the ideas, they
                have been very helpful!

                Geoff

                --- In nmrpipe@yahoogroups.com, Ryan McKay <ryan@n...> wrote:
                > If you skip the first HT does the second HT then cause poor results

                > (i.e. is it always the first HT no matter where it is performed
                that is
                > causing errors) or is it one particular dimension?
                >
                > Have you checked the raw FIDs along that dimension for a bad point?

                > For example in one 3D-13C-NOESY I like to use the first two points
                of
                > the 13C dimension can be distorted if the sweep widths are too
                wide.
                > Back predicting these two points or using a 1/2 dwell delay prior
                to
                > acquisition removes the fault and results in the proper spectra.
                >
                > When you say that it results in poor spectra are you referring to
                > signal to noise, peak shape distortions, water suppression etc.?
                >
                > Thanks,
                > Ryan
                >
                >
                >
                >
                > On Jun 14, 2004, at 11:40 AM, Geoff Armstrong wrote:
                >
                > Thanks for the responses, the suggestions have been helpful.€  '    
                But I
                > have not yet been able
                > to get satisfactory results.€  '      Just doing the LP of the
                untransformed
                > data as per Ryan's
                > suggestion worked but the results were very poor.€  '    
                Analyzing
                the
                > individual files produced
                > by each step of the script suggests that the problem occurs during
                > the first inverse
                > processing step as the result is incorrect if I inverse process
                and
                > then reprocess (without
                > any LP) any of the indirect dimensions.€  '     Could it be that
                the
                Hibert
                > Transform is not
                > working for my data?€  '     I've been through the options for HT,
                but it
                > doesn't seem like any of
                > them apply.€  '     My data is constant time, but does not have a
                1/2
                dwell
                > delay.€  '     Any ideas?€  '    
                > Thanks in advance!
                >
                > Geoff
                >
                > --- In nmrpipe@yahoogroups.com, "David A.€  '     Horita"
                <dhorita@w...>
                > wrote:
                > > Geoff,
                > > You may also have a problem with the last step being -in -a
                -out -y
                > without the inPlace
                > because this arrangement might over-write itself.€  '     In
                addition
                to
                > Ryan's suggestion, use
                > different output names from the input files so you can see where
                the
                > processing fails.
                > >€  '    
                > > Regards,
                > > David
                > >
                > > €  '        
                > >
                > > €  '         xyz2pipe -in lp/test%02d%03d.dat -a
                -verb€  '             
                \
                > > €  '         | nmrPipe€  '     -fn HT€  '               
                €                         \
                > > €  '         | nmrPipe€  '     -fn PS€  '     -p0 0.0 -p1
                0.0€  '              
                -inv€  '     \
                > > €  '         | nmrPipe€  '     -fn FT€  '   
                                            
                -inv€  '     \
                > > €  '         | nmrPipe€  '     -fn ZF€  '   
                                            
                -inv€  '     \
                > > €  '         | nmrPipe€  '     -fn SP€  '     -off 0.5 -end
                0.98 -pow 2€  '    
                -inv€  '     \
                > > €  '         | nmrPipe€  '     -fn LP€  '     -ps0-0€  '    
                €                            \
                > > €  '         | nmrPipe€  '     -fn SP€  '     -off 0.5 -end
                0.98 -pow
                2€  '           \
                > > €  '         | nmrPipe€  '     -fn ZF€  '     -auto€  '     
                €                            \
                > > €  '         | nmrPipe€  '     -fn FT€  '               
                €                         \
                > > €  '         | nmrPipe€  '     -fn PS€  '     -p0 0 -di€  ' 
                €                            \
                > > €  '         | pipe2xyz -out lp/test%02d%03d.dat -y -ov
                > > €  '        
                > > €  '         €  '    
                > > €  '        
                > >
                > > -----------------------------
                > > David A. Horita, Ph.D.
                > > Department of Biochemistry
                > > Wake Forest University School of Medicine
                > > Winston-Salem, NC 27157-1016
                > > Tel: 336 713-4194
                > > Fax: 336 716-7671
                > > email:€  '     dhorita@w... <mailto:dhorita@w...>
                > > web:€  '     http://www.wfubmc.edu/biochem/faculty/Horita/
                > <http://www.wfubmc.edu/
                > biochem/faculty/Horita/>
                >
                >
                >
                >
                >
                > Yahoo! Groups Sponsor
                >
                > ADVERTISEMENT
                > <yhoo0504_testa_300250a052604.gif>
                > <l.gif>
                >
                > Yahoo! Groups Links
                >
                > €  '   '²µ To visit your group on the web, go to:
                > http://groups.yahoo.com/group/nmrpipe/
                > €  '    
                > €  '   '²µ To unsubscribe from this group, send an email to:
                > nmrpipe-unsubscribe@yahoogroups.com
                > €  '    
                > €  '   '²µ Your use of Yahoo! Groups is subject to the Yahoo!
                Terms
                of
                > Service.
                >
                >
                >
                >
                >
                > ------------------------------------------------
                ------------------------
                > ------------
                > Ryan T. McKay, Ph.D. - Scientific Director
                > NANUC - Canada's National High Field NMR Centre
                > Room 105 Email: ryan@n...
                > NANUC Bld. Phone: (780) 492-2700
                > University of Alberta Fax: (780) 492-9174
                > Edmonton, Alberta Cell: (780) 920-8871
                > Canada T6G 2E1 AIM: ryanmckaynmr
                > http://www.nanuc.ca
              Your message has been successfully submitted and would be delivered to recipients shortly.