Experiment : Spectra

Curate Spectrum Parameters

This popup window allows the user to view and edit information that is associated with spectra. In this regard a spectrum is a record of the NMR data that was obtained. Parameters that relate to what was done experimentally to get the data are specified elsewhere; experiment level information is set via the main Experiments popup window. Many spectrum level parameters in Analysis relate to how the spectrum data files are interpreted and displayed graphically, i.e. as contours. Data that derive from the analysis of spectra, like peak lists and assignments, are managed elsewhere in specialised displays.

The layout of this popup window involves several tab panels that sub-divide the spectrum settings into related groups. The functionality within each tabbed sub-division is largely independent, although a spectrum entity selected in any of the tables or pulldown menus is subject to the [Delete Spectrum] function.

Main “Spectra” tab

This table presents an overview of all of the spectra that have been loaded into the CCPN project, what their dimensions are and how they relate to experiments and peak lists. Some of the parameters may be edited by double-clicking on a cell in the table, thus for example the user may change the name and set the active peak list (the one that is used for peak peaking and selection). In the main however, this table usually serves simply to illustrate what is present, which experiment each spectrum belongs to and what its NMR dimensions are.

Display Options

This table is used to control how spectra are visualised within the spectrum window displays of Analysis. Specifically, the user can: change the colours of the displayed contours and 1D slice traces; get access to the contour level settings, via the Spectrum Contour Levels tool; access the Spectrum Contour Files setup; set keyboard toggles; and administer other graphical elements like bounding box lines, peak fonts and peak pointer lines. It should be noted that the colours for peak text and peak symbols are set for individual peak lists, so that different peak lists within the same spectrum may be distinguished. Accordingly these parameters are set via the main Peak Lists option.

The “Rank” value of a spectrum dictates the order in which spectra are drawn on screen and presented in the toggle options at the top of spectrum windows. The rank may be set on an individual basis or following the order of spectra in the table using [Set Ranks From Order]; this is especially useful if the table is first sorted in a particular way (by clicking on a table heading).

Referencing

This table is used to control how the points in a spectrum data file are interpreted in terms of NMR frequency, on each of their axes. Usually, this means to know how positions in the spectrum data, stored as an array of intensity data points, relate to ppm units and Hz frequency values of the spectrometer. Such referencing occurs by specifying how the width of a spectrum dimension in data points relates to its width in frequency units, defining the scaling factor, and how the two scales are anchored together; by giving a reference data point and its corresponding a frequency/ppm value.

The spectrum referencing values are normally extracted from the spectrum header or parameter file when the spectrum was first loaded. Often these values do not need to be altered. However, the user may still adjust all of the parameters after load time, with the sole exception of the dimensions isotope type. Changes to spectrum referencing may be made after loading to correct for mistakes in the initial data interpretation and to tweak spectra so that they give a better overlay of signals. Note that the table columns ‘Spectrometer Frequency’, ‘Minimum Aliased Frequency’, ‘Maximum aliased frequency’, and ‘Sub Dim Name’ (normally hidden) apply to the experiment, and will are shared between all spectra in the same experiment.

Peaks that are picked within spectra have positions defined in terms of spectrum data points, i.e. normally indicating an extremum in intensity. The referencing values of a spectrum’s dimensions are used to calculate what the position of a peak is in ppm or Hz units. Accordingly, if spectrum referencing values are changed then peak ppm locations will naturally move. If peaks should not move with the spectrum data when referencing is changed, to instead stay at the same ppm values, then the user can select “ppm” in the “reference changes keep constant peak:” options.

Some less common experiments have more than one type of signal (shift) on each axis, and need a reference line to describe each experiment type. One obvious case is projection experiments, where signals in one dimension are linear combinations of frequencies for different nuclei. Another would be the 15N/13C HSQC-NOESY, where 15N and 13C are measured in parallel on one axis. The special options for this situation are turned on by selecting ‘Use reduced dimensionality options’. The individual reference lines are called ‘Sub-dimensions’; they are shown explicitly when there is more than one. They can be added, removed, and edited individually. The Sub-dimensions set for the first Spectrum serve to (re)define those for the Experiment; once there is more than one Spectrum in an Experiment it is not possible to add extra Sub-dimensions. Projection spectra can only be read correctly from some formats (Bruker and USF3) and even these cases may require editing. The special ‘Sub Dim Name’ column gives a name that is used by some external programs (e.g. Prodecomp) to identify a given signal type across different experiments. The special ‘Scaling Factors’ column gives the scaling coefficient that applies to a particular signal in this spectrum. For instance scaling factors CO:1; CA:0; CB:-2 would denote a signal of the form wN-2wCB. There can be more than one scaling factor where different types of peaks are possible, e.g. wN OR wC. The system has only been tested for 2D projection spectra.

The minimum and maximum aliased frequency values for a spectrum dimension, as displayed in the last columns of the table, may be set to cut off the spectrum contour display at a particular value or extend the contour display. Contour displays that are extended beyond the normal sweep width (ppm range) of the spectrum do so by tiling; a copy of the contours (not mirror image) is made and placed next to the normal, fundamental region. This mechanism is used to cope with aliased resonances that appear (as ‘ghost’ peak images) within the main spectrum width in places that are not their true ppm values; the resonance is really outside the sweep width of the spectrum. Making tiled copies of the contours allows the spectrum to extend and cover the real ppm values of aliased peaks. Aliased peaks may be moved to their true ppm locations or picked at that location in the first instance by using the appropriate tile region.

Tolerances

This section is used to control the weightings and tolerances that apply to individual spectrum dimensions during the resonance assignment process. There are two kinds of value that can be set. One is the “Assignment Tolerance”, which states how close (typically in ppm) chemical shift values of resonances match to peak peak locations; this is an upper limit such that resonances with shifts outside the tolerance are not considered as assignment possibilities. The second thing is the “Shift Weighting” that specifies how much influence resonance assignments have on the averaged chemical shift values. Both of these parameters affect individual spectrum dimensions because each axis has is its own resonance line shape and data resolution.

The normal chemical shift calculation within Analysis calculates values as the weighted mean over all of the peak positions to which a resonance is assigned. If a dimension has a shift weighting of 2.0 but others have 1.0 then any resonance assignment in the former will have twice as much influence on the shift mean, as if it were assigned to two peak dimensions rather than one. By default all weights are set to 1.0, but the user can increase this if a spectrum dimension is particularly precise (narrow peaks) or decrease it if peaks are broad or if resonance assignments are highly ambiguous; for speculative NOE assignments the user usually doesn’t want the shift values to change according to those assignments.

The assignment tolerance and shift weighting values are set for the dimensions of a selected spectrum in the upper table. The lower “Spectrum Overview” table shows these values across all spectra in the project, so that the user can see the overall situation and range of values. Although tolerances are not set in the lower table on an individual basis, the assignment tolerances and shift weights can be spread from one spectrum (the last selected) to any others selected in the table (using left-click + <Crtl> or <Shift), where the isotopes of dimensions match.

File Details

This tab is used to indicate the location of a spectrum’s binary data file and how the binary data within is read. In essence the binary values of the file need to be interpreted as spectrum intensity values, but in order to do this the system has to know how the numbers are stored. This involves knowing how numbers are represented per se; how many bytes a number uses, whether numbers are integers or floating point and what orientation the bits are in (big or little endian). Also, most spectrum data formats involve binary blocks; chunks of data organised into a grid across all the dimensions. The benefit of using blocks comes when looking at large data sets, e.g. using 3D and higher dimensionality data, because it means that when only looking at a sub-region of a spectrum, like a single plane of a 3D, only part of the spectrum data needs to be read; most of the spectrum data that isn’t displayed does not need to be loaded. Naturally, the system needs to know what the sizes of the data block along each dimension are. These are displayed in the lower table and may be adjusted if there are any mistakes, which may have been inherited from the spectrum processing.

The “File name” field shows the present location of the binary data. If this file is missing the spectrum will still be represented in Analysis, but the operations which rely of having intensity data, like picking peaks, cannot be performed. It is notable that the binary data file for a spectrum can be changed at any time, either by selecting a different file name (typing directly or using [Choose File]) or by directly replacing the binary data inside the file. If the binary data or any of the parameters that influence its interpretation has changed, then the user can click on [Refresh Spectrum] to see the effect if the changes.

Data Locations

This last table is used to show the locations of all of the spectrum data files used by the whole project. These locations need to be changed if the spectrum data is moved in disk, otherwise the system will not know where to look. Each data location is broken down into two components; an “absolute” directory path and a “relative” path, which contains the file name. Both of these components may be changed.

The absolute part of the location is from the root of the file system (the all-encompassing directory) and is often shared by several spectrum files; it is typical to store spectra in one repository. For convenience, changing the absolute part of the path for one spectrum will usually change the value for all the other spectra that use the same directory. Also, if multiple files are moved to the same directory then [Propagate Absolute Path] can be used to state that several files are in the same location.

The relative part of the spectrum data path is individual to the spectrum and contains the name of the data file and any extra directory location, relative to the absolute part of the path. Which parts of the spectrum file path is considered to be absolute or relative may be changed using the “Shift Directory...” buttons. For example an absolute : relative split of “/data/spectra/” : “protein/123.spec” can be changed to “/data/spectra/protein” : “123.spec”, so that several spectra can use the absolute part.

Caveats & Tips

The spectrum chemical shift weighting used for synthetic peak lists is set to 0.0001 of the the regular weighting for the spectrum dimension set via this popup.

The fonts used for the printing of spectra (e.g. via PostScript of PDF) may be set separately to the main spectrum font specification, which is used within Analysis spectrum displays.

Main Panel

button Delete Spectrum: Delete selected spectrum or spectra

button Clone: Clone popup window

button Help: Show popup help document

button Close: Close popup

Spectra

Basic information about spectra, their experiments, peak lists & dimensions

Table 1
# Row number
Experiment The name of the experiment to which the spectrum belongs (Editable)
Spectrum A short name for the spectrum, often used with experiment name for unique identification “experiment:spectrum”. e.g. “HSQC:115” (Editable)
Dimensions Number of dimensions/axes in the spectrum (can be fewer than in the experiment) and their isotope types
Active Peak List The active peak list, for example the peak list into which picked peaks are put (Editable)
Num Peak Lists Number of separate peak lists within the spectrum
Scale Spectrum scale relative to stored data points, e.g. for drawing (Editable)
Noise Level Estimated intensity level of spectrum noise (Editable)
Data Type Whether fully processed, FID or part-processed
Simulated? Whether the spectrum is simulated
Data File Location of the data file relative to the data directory

Display Options

How spectra are displayed on screen in terms of colours & contours etc.

Table 2
# Row number
Spectrum Identifying name Experiment:Spectrum
Positive Colours Colour scheme for positive contours (can be a single colour) (Editable)
Negative Colours Colour scheme for negative contours (can be a single colour) (Editable)
Slice Colour Colour of 1D slice and trace lines (Editable)
Box Visible? Whether fundamental region (sweep width bounding) box should be drawn (Editable)
Shortcut Shortcut key for toggling spectrum on/off (Editable)
Rank Determines order spectra drawn in (lowest appears on top) (Editable)
Positive Contours Positive contour levels, relative to global scale (Editable)
Negative Contours Negative contour levels, relative to global scale (Editable)
Use Contour Files Whether pre-calculated contours should be used (Editable)
Peak Font The typeface used for peak annotation in spectrum window displays (Editable)
Peak Pointer? Whether the peak pointer (a triangle/line from annotation to centre), if it exists, should be drawn (Editable)

button Contour Levels: Open popup for detailed setting of contour levels

button Contour Files: Open popup for administering pre-calculated contour files

button Propagate Contours: Propagate contour levels from row selected last to others

button Set Ranks From Order: Set spectrum rank based on current order in table (e.g. after sorting on a column)

Referencing

Information for interpreting data points in terms of PPM & Hz

pulldown Spectrum: Select the experiment:spectrum to adjust

check Use reduced dimensionality options: Used for reduced dimensionality or projected experiments when you have sub-dimensions

radio point: If you change the referencing then peak position in points stays the same

radio ppm: If you change the referencing then peak position in ppm stays the same

Table 3
Dim Spectrum dimension or sub dimension
Isotope Isotope being recorded
Spectrometer frequency (MHz) Carrier frequency for the particular isotope, in MHz (Editable)
Spectral width (ppm) Actual spectral width, so possibly after region truncated, in ppm
Spectral width (Hz) Actual spectral width, so possibly after region truncated, in Hz (Editable)
Reference ppm Ppm value at chosen reference point (Editable)
Reference point Point value where dimension has chosen reference ppm (Editable)
Orig. number of points Original number of points, so after Fourier transform and before any truncation of the region (Editable)
Point offset Offset of region relative to original region after Fourier transform, so 0 if the region has not been truncated (Editable)
Minimum aliased frequency (ppm) Minimum frequency, in ppm, of where any signal is expected, if not set then uses minimum frequency of fundamental region (Editable)
Maximum aliased frequency (ppm) Maximum frequency, in ppm, of where any signal is expected, if not set then uses maximum frequency of fundamental region (Editable)

button Add Sub-dimension Copy: Add a new sub dimension (e.g. reduced dimensionality component) by copying the selected sub-dimension

button Remove Sub-dimension: Remove selected sub-dimension

Tolerances

Per-dimension assignment tolerances & chemical shift calculation contributions

pulldown Spectrum: Documentation missing

Table 4
Dimension Spectrum dimension
Isotope Isotope recorded in this dimension
Assignment Tolerance When assigning, the maximum difference between a peak dimension position and a known chemical shift value (Editable)
Shift Weighting Contribution, in this dimension, of peak positions to the averaged, weighed chemical shift values (Editable)

Spectrum Overview

Table 5
# Row number
Spectrum Experiment name : Spectrum name
Dim Assign Tolerances Peak vs shift assignment tolerances in the various dimensions, as set in the Spectrum table
Dim Shift Weights Chemical shift weighting contribution, in the various dimensions, as set in the Spectrum table

button Propagate Assignment Tolerances: Propagate peak vs shift assignment tolerance from last selected spectrum to other selected spectra in Spectrum Overview table

button Propagate Shift Weights: Propagate chemical shift weighting contribution from last selected spectrum to other selected spectra in Spectrum Overview table

File Details

Location of spectrum data files and how data is stored in blocks & points

pulldown Spectrum: Select the experiment:spectrum to operate on

Data File Details

entry File name: Full path to the file containing the spectrum data (typically binary)

check Is data big endian?: Whether binary numbers in data are big endian (e.g. Mac PowerPC, SGI) or little endian (e.g. Intel), which depends on how and where data was collected and processed

pulldown Bytes per word: Bytes per data point (normally 4)

pulldown Data number type: Whether data is floating point or integer (normally floating point)

int File header size: The header size in bytes, which depends on the file format

button Choose File: Select a new spectrum data file using a dialog, rather than typing it above

button Refresh Spectrum: Save the information as chosen and refresh spectrum display

Table 6
Dimension Dimension in data file
Number Points Number of data points in this dimension (Editable)
Block Size Size of blocks into which data points are partitioned for efficient reading (Editable)

Data Locations

A table of all spectrum data locations for en masse manipulation

Table 7
# Row number
Absolute Path Location of the absolute path directory, that may contain multiple spectrum files (Editable)
Relative Path and File Location of spectrum file relative to its absolute path directory (Editable)
Spectrum / FID / Image Identity of the spectrum as experiment:spectrum

button Propagate Absolute Path: Copy absolute path from last row selected to other selected spectra

button Shift Directory To Absolute Path: Shift the top directory from the relative path to the absolute path

button Shift Directory To Relative Path: Shift the bottom directory from the absolute path to the relative path

Table Of Contents

Previous topic

Experiment : Open Spectra

Next topic

Experiments : Experiments

This Page