Result File Guide

This Page lists every property of an entry in our result files with a description of what it represents, how it was calculated and how you should interpret it. (WIP)

All masses are in Dalton, assuming a charge of 1, and all retention times are in seconds! The intensity is in the unit supplied by the mass spectrometer. Scores do not have a unit.

For the file "example.mzML":

First, the polarity is checked. The polarity is added as "positive" or "negative" to the output file name. If polarity switching is used, two result files are generated.

example_positive_centroids.csv

Every centroid is a regression determined using the qPeaksalgorithm.

• ID: Integer number counting upwards by scans and m/z within scans.
• mz: m/z value for the centroid, calculated as the apex position of the regression.
• mzUncertainty: Prediction interval for the distribution of m/z that was centroided.
• scanNumber: Which scan of the mass spectrometer this centroid exists in.
• retentionTime: Retention time, including the void time (the first scan is at t = 0).
• area: Area under the gaussian curve described by the intensity profile.
• areaUncertainty:
• height: Maximum of the regression fitted to the centroid.
• heightUncertainty
• degreesOfFreedom: Degrees of freedom of the regression. They are calculated as the number of points in the centroid - 4.
• dqsCen: Data quality score of the centroid. Calculated as the error function of areaUncertainty divided by area.

example_positive_bins.csv

Every bin is a group of centroids that was determined to have very similar m/z values over a range of continuous scans, with a tolerance for gaps no greater than 3. Since the output of qCentroids is used, only those properties added by the binning step are explained.

• binID: Integer number counting upwards by bin generation point. It does not hold chemical information.
• DQSB: Quality score of the centroid in the bin. It is calculated as the ratio of the mean distance of a centroid to all other bin members within four scans and the smallest difference to a centroid that is not part of any bin. It ranges from -1 to 1, with values > 0 meaning that this point is more similar to the bin than its surroundings.

example_positive_features.csv

Every feature is a regression over a bin.

• ID: Integer number counting upwards by binID.
• binID: The bin in which this feature was found.
• binIdxStart: First point within the bin that is part of the region over which the feature was determined.
• binIdxEnd: Last point within the bin that is part of the region over which the feature was determined.
• mz: m/z value for the feature, calculated as the weighted mean by intensity.
• mzUncertainty:
• retentionTime: Apex position of the regression in RT.
• retentionTimeUncertainty:
• lowestRetentionTime: RT at the point binIdxStart.
• highestRetentionTime: RT at the point binIdxEnd.
• area: Area under the gaussian curve described by the intensity profile.
• areaUncertainty:
• height: Maximum of the regression fitted to the bin.
• heightUncertainty
• dqsCen: Weighed mean by intensity of all values for dqsCen within the regression window.
• dqsBin: Weighed mean by intensity of all values for dqsBin within the regression window.
• dqsPeak: Data quality score of the Feature. Calculated as the error function of areaUncertainty divided by area.

---