﻿ Peak Picking details - EN

# Online Software Help Manual

 Contents
 Display Legacy Contents

# Peak Picking details

The Peak Picking function of the software automatically detects peak maxima or minima in a 2D data object. Either discrete or equidistant data objects can be used. Peaks can also be added and removed manually.

In automatic peak detection, only peaks with a user defined minimum intensity relative to the data baseline and a minimum average peak width will be considered. These parameters must be adjusted properly by the user before automatic calculation is applied.
From the first derivative of the data object, maxima will be detected using the Savitzky-Golay derivative function. Shoulders and/or overlapping of adjacent peaks might be also analyzed using the second derivative of the data object. Some parameters can be adjusted by the user to optimize peak finding conditions.

When using manual peak finding, the user may freely define new peaks at any data position. After adding a new peak its start, end and maximum position can be modified.

Resulting peak information will be shown either in a peak table or as peak markers directly in the data view or both. Display options can be customized in the 2D Preferences dialog. The peak table is shown on a separate tab below the data view. It will be displayed on demand.

## Peak table

The peak table holds several columns with information regarding position of a peak within a 2D data object. Available columns strongly depend on the data type.

Usually, the following columns are available in a peak table:

### Number

Each peak in the collection possesses a unique, concurrent number starting from the lowest x-axis value to the highest x-axis value.

### Begin

This value indicates the starting point of a peak in x-axis units.

### Max

This value indicates the position of the peak maximum/minimum in x-axis units.

### End

This value indicates the ending point of a peak in x-axis units.

### Width

This value indicates the difference between ending point and starting point in x-axis units.

### Height

This value indicates the relative intensity at the peak maximum in y-axis units. The relative intensity at the peak maximum will be calculated as difference between the absolute intensity and the baseline intensity at the maximum.

### Absolute Height

This value indicates the absolute intensity at the peak maximum in y-axis units. The absolute intensity at the peak maximum will be calculated as difference between the absolute intensity and zero.

### Peak Area

This value indicates the peak area between the graph and the baseline, lasting from the starting point to the ending point of a peak.

### Center of Gravity

This value shows the Center of Gravity of the peak.

## Peak table display options

Visibility of peak table columns and order of appearance can be customized. Whenever a peak table is shown in the peak table tab customization is carried out as described in the following:

 Results from Peak Evaluation are also visible here! If any peak evaluation methods are defined for the current data type, the results are available in additional peak table columns. These columns can be shown or hidden here. Please refer to the "Peak Evaluation" section for details.
1. Click the Right mouse button on the peak table area.

2. From the dialog, select preferred columns to be displayed.

3. Change their order of appearance according to you needs.

4. Click the OK button to apply settings

## Peak Markers

Besides the peak table, peak markers are shown in the data view area on top of the corresponding peak. Each peak marker consists of start, end and maximum/minimum tick marks as well as a peak label. The peak label shows the actual position either on the x-axis, y-axis or both. Peak markers can be customized in the 2D Preferences dialog.

## Find peaks automatically

The following excerpt from the N,N,2,4-Tetra-Methyl-Aniline IR spectrum is submitted to peak picking in the following. The spectrum excerpt looks like this:

The minimum peak height must be adjusted properly to give positive peak finding results. If the value is too large, smaller peaks will not be detected anymore. It should be adjusted close to the noise level of the spectrum to gain optimal results.

The noise level including some minor baseline drifts up and down is estimated to be 0.02 y-axis units. So the minimum peak height parameter should be set to this value.

 How to get the optimum peak height level? Start with 0 as minimum peak height and minimum peak width. Then increase the minimum peak height value in small y-axis fractions to get all the peaks you like to see or even some more.

The peak width must be properly adjusted. If it is set too small, all kinds of noise or spikes will be detected and false peak results will be produced. If the value is set too large, narrow peaks might be missed and will be kicked out of the peak list. So an average peak with for the current data type should be entered here.

 How to get the optimum peak width level? Start with the previously adjusted minimum peak height level and enter the width of the broadest peak in your spectrum. Adjust this value as minimum peak width. Then decrease the value in small x-axis fractions to slowly increase the amount of detected peaks to get as many peaks as you like to see.

### Finding peaks without adjacent peak grouping

This option is used to separate pre-selected peaks (see parameter adjustment above) that might overlap to just one peak.

Peak finding without adjacent peak grouping is shown in the following figure:

The following parameter settings are applied:

• Group adjacent Peaks = No

• Minimum Peak Height = 0.02

• Minimum Peak Width = 0

All the center peaks are strongly overlapping, but the peak grouping is deactivated for finding peaks. In this case, the baseline strongly follows the graph slope. It moves up to the small peak in the middle and down again. The large peak in the middle is split into a total of three peaks.

### Peak picking with adjacent peak grouping

In contrast to the figure above, the following figure shows the same peak detection with adjacent peak grouping.

The following parameter settings are applied:

• Group adjacent Peaks = Yes

• Minimum Peak Height = 0.02

• Minimum Peak Width = 0

The baseline does not strongly follow the graph slope in this case. Adjacent or overlapping peaks, which do not return to the overall spectrum baseline level, will be cut off by a vertical line. Area and peak height calculation will be adapted to the baseline accordingly.

1. Move the mouse pointer close and underneath the position of the graph where you like to add a peak.

2. Hold down the CTRL-key and click the Left mouse button to add a new peak.

3. Use the red tracker boxes to adjust the bounds of the manually added peak.

## Edit peaks manually

Peaks can be edited manually either by modification of start, end or maximum values within the peak table or graphically. Modification of peaks is possible as long as the Peak find tool is activated in the mathematics tab.

### Modifying peaks in the peak table

Modification of a peak within a peak table is done as described in the following:

1. Browse the peak table or just click the peak to be modified in the data view with the Left mouse button.

2. Click the Left mouse button to enter the field you like to change.

3. Enter a new valid value.

 Why is the entered value not valid? The new entered value for start and end of a peak must be chosen in that way, that the peak area underneath the peak is not parted by the graph. In other words, the baseline of the peak must not cross the graph. The value will be adjusted automatically to the nearest valid data point to your actually selected position. Secondly, the position of the maximum/minimum must be within the limits of start and end peak marker.

1. Press the Return-key to apply changes.

### Graphical modification of peaks

As long as the peak find tool is activated in the mathematics tab, each peak shows tracker icons (red squares) at start, end and maximum/minimum position of a peak. Modification of the peak is performed as described in the following:

1. Move the mouse pointer close to a tracker icon.

2. Press and hold the Left mouse button.

3. Move the mouse along the graph slope to the new destination position.

4. Release the Left mouse button.

 Why is my selection not applied? The new start or end position of a peak must be chosen in that way, that the peak area underneath the peak is not parted by the graph. In other words, the baseline of the peak must not cross the graph. The position will be adjusted automatically to the nearest valid data point to your actually selected position. Secondly, the position of the maximum/minimum must be within the limits of start and end peak marker.

## Removing peaks

Single peaks or the whole peak table can be removed from the current data object.

### Removing a peak table

1. In the mathematics tab, select the Find Peaks operation.

2. Press the Clear Table button.

### Remove a single peak

1. Select the peak to be removed either in the data view or the peak table by clicking with the Left mouse button onto it.

2. Press the DEL-key to remove it.