Identifying outliers in DLS measurements

In some cases, PR.Panta Control can single out problematic dynamic light scattering acquisitions based on intensity variations or average scattering intensity and exclude them from the calculations of average values in size analysis measurements or the diffusion interaction parameter kD. The software makes use of quality checks to flag outliers, which can be overridden by the user.

 

If the average scattering intensity calculated from a DLS acquisition (sum of detected signals over time) exceeds 109 counts/s, the detector will be saturated. If this occurs, the resulting output parameters will be incorrect, and the corresponding acquisition will be automatically marked as an outlier.

 

If single acquisitions of repeated measurements of the same capillary during size analysis measurements display an average scattering intensity that is very different from the remaining acquisitions from that capillary, they are also flagged as outliers. PR.Panta Control will flag outliers due to these intensity variations if the scattering intensity is more than 20% higher than the average scattering intensity of the passed acquisitions of a capillary.

 

Acquisitions automatically or manually flagged as outliers are highlighted in orange in the plots displaying multiple acquisitions:

 

OutlierPlot_657x480.png

 

Figure: Violin plot showing the size distribution and the hydrodynamic radius of a size analysis measurement. The rH of the cumulant fit is displayed as a purple dot, while the results of the size distribution fit are shown as blue lines. The outlier is marked in orange. The violin plot is a size distribution mirrored on the radius axis, which contains an overlay of all acquisitions of a capillary, therefore each line/dot depicts a single DLS acquisition.

 

How many acquisitions passed the outlier rating and are therefore used for calculation of average values of the size analysis data is visible in the tabulated results. If less than 80% of the acquisitions of a capillary pass the rating, the warning symbol will turn orange and the corresponding table cell is highlighted:

 

OutlierTable.png

 

Outliers are also automatically marked in thermal unfolding measurements if the average scattering intensity exceeds the upper limit.

 

Was this article helpful?