The Med Res. Feature Detector node uses the sliding window and the kDecon algorithms to perform spectral deconvolution and measure all of the deconvoluted features and their quantitation traces. The parameters used in this node are essential to the label-free quantitation workflows. Set the parameters according to the data being analyzed.

The Sliding Window Averaging Width RT parameter establishes the RT for the sliding window average. This value should match the full width at half max of a representative peak from the analyzed data. Thermo Fisher Scientific recommends measuring the peak width of an average feature in your data set. Do not choose the most abundant peak or one that is at noise level.

The following table lists the node parameters.

Table ProSightPD Med Res. Feature Detector node

Parameter

Description

1. kDecon

Precursor Maximum Mass

Specifies the maximum precursor mass to be considered by kDecon.

Precursor Minimum Mass

Specifies the minimum precursor mass to be considered by kDecon.

Number of Results

The maximum number of kDecon results to return for each averaged scan. Results are filtered by S/N prior to being returned.

Number of Iterations

kDecon can be run multiple times. After each iteration, charge state distributions from each mass are removed from the spectrum to limit false positives during subsequent iterations.

S/N Result Cutoff

The geometric mean signal-to-noise ratio S/N is calculated from the individual S/N of each charge state in the charge state distribution of a detected mass.

Peak Tolerance in PPM

Peak Tolerance PPM

2. Sliding Window Parameters

Averaging Width RT

Specifies the retention time, or the width, of the sliding window, in minutes. Reducing this value improves time resolution but reduces execution speed and possibly sensitivity. Increasing this value increases execution speed but reduces time resolution and might increase sensitivity. You can achieve the best results when the window width is between one-quarter and twice the width of the characteristic peaks in the spectrum. For most usage, the optimum value might be half the width of the characteristic peaks. For example, if those peaks have a width of one minute, the optimum width would be 0.5 minutes.

Offset Type

Specifies the offset between successive sliding windows as a number of scans or as a percentage value. You can select from these two options: scan-based offset or percentage of averaging width RT offset.

Offset Scan(s)

This mode offsets each window from its predecessor by the user-specified number of scans. An offset of n means that each window begins n scans after the beginning of its predecessor.

Offset Percentage

This mode offsets each window from its predecessor by the user-specified percentage of the window width. An offset of 30% means that each window begins 30% after the beginning of its predecessor and overlaps the last 70%. An offset of 100% means that successive windows are adjacent with no overlap.

Merge Tolerance (ppm)

Determines how close two components in successive sliding windows must be in mass (ppm) for the application to identify them as a single component. A value that is too high might result in merging components that should remain separate. A value that is too low might result in false positives when components remain separate that should have merged.

Minimum Number of Charge States Observed

Minimum required number of observed charge states for data to be analyzed by KDecon. For top-down data, the recommended minimum is 3. Some very low-abundance species might only have 2 observable (greater than noise) peaks. However, setting this value to less than 3 will allow noise through and result in false positives.

Minimum Number of Sliding Window Detections

Specifies the minimum number of sliding window intervals that a component must appear in for the application to consider the component valid. A value that is too low might allow noise peaks to appear as false positives. A value that is too high might result in legitimate components being discarded. Set this parameter to a value large enough to exclude results that are implausibly narrow in retention time but small enough to include results of realistic duration. Values in the range of 3 through 8 generally give good results. A good approach is to use whichever is larger: 3 or the minimum number of windows that can fit into a peak.

Biggest Gap Type

Specifies the allowed separation between successive individual members of a merged component identified by the Sliding Window Deconvolution algorithm. You can select from these two options: Scans or Retention Time.

Biggest Gap Scans

The maximum allowed separation in scans between two successive individual members of a merged component identified by the Sliding Window Deconvolution algorithm. If the separation exceeds this value, the component will be divided into two components separated. As with the Merge Tolerance, too high a value can result in components that should have remained distinct being merged, while too low a value can result in false positives when components that should have been merged remain distinct. This parameter should be comparable too or slightly less than the expected separation in scans between distinct components with the same mass.

Biggest Gap Retention Time

The maximum allowed separation in retention time between two successive individual members of a merged component identified by the Sliding Window Deconvolution algorithm. If the separation exceeds this value, the component will be divided into two components separated by the observed gap in retention time. As with the Merge Tolerance, too high a value can result in components that should have remained distinct being merged, while too low a value can result in false positives when components that should have been merged remain distinct. This parameter should be comparable too or slightly less than the expected separation in retention time between distinct components with the same mass.

3. Feature Grouping

Mass Tolerance

Mass tolerance used for calculating mass traces.

RT Threshold

Retention time limits used for calculating features.

4. Feature/Trace Connection

Trace Mass Tolerance

Specifies the trace tolerance.

Number of Smoothing Points

Specifies the number of points to average. As the number increases, the data become smoother resulting in a loss of fine features. The recommended number of points is 3.

Time Range (min)

The allowable time difference between a trace and its linked feature. The recommended setting is 1 minute.

Trace Smoothing Type

Specifies the type of smoothing to perform: Gaussian, Moving Average, or None. The recommended setting is None. This should be used only in cases where unsmoothed data is not acceptable.

5. Multithreading Options

CPU Usage

Amount of CPU to direct towards data processing.