Frequently asked questions about BioAnalyte's Regatta
Questions
General
Multivariate analysis and algorithms
- How much information from the trawl files does the HCA clustering consider? Does it account for mass, RT and intensity?
- What is dynamic binning and how is it related to alignment?
- I have two groups of data, Control and Test, which cluster nicely. Why is the length of the AND(Control Samples) greater than the length of the NODE(All Control Samples) in my cluster? Aren't they supposed to report the same numbers?
Answers
General
Where is the 1D pseudo-gel view?
Select the list(s) in the Active Workspace, right click and select "1D Pseudo Gel View."
Is it possible to change list labels?
Yes. Turning off the display (checkbox=OFF). Then, click on the list label and edit. When you turn the display on again, the new name will propagate through the GUI.
It is not possible to edit the list type.
Is it possible to change the colors in the stick plot and 2D map displays?
No, not in the current version.
Is it possible to change the order or stacking of the displayed items?
No, not in the current version.
Multivariate Analysis and Algorithms
How much information from the trawl files does the HCA clustering consider? Does it account for mass, RT and intensity?
Currently, the standard Presence/Absence HCA tool uses both mass and retention time to determine the clusters. The intensity weighted overlap metric does make use of intensities. We've added a Feature Request to make it sensitive to one or the other in addition to both.
What is dynamic binning and how is it related to alignment?
Dynamic binning is the process of taking several lists of analytical data with errors (apples and oranges) and rationalizing the measurements to satisfy the Transitive Property of Equality, thereby making it possible to compare apples to apples and oranges to oranges. It is a subtle and complicated process that makes data analysis independent of the order of analysis, which is a common artifact in analytical list analysis. Most analytical chemists believe that a specific retention-time-only alignment must precede dynamic binning, we simply treat retention time measurements as having large error bars. We haven't yet seen data that suggests that the computational effort expended in shift retention time by a few seconds one way or the other is worth the cost in doing so. By having error bars large enough to encompass the jittery performance of the chromatography, we subsume the alignment process within the dynamic binning process. The result is much, much faster processing, albeit at the cost of precision in retention time. Fortunately, most analytical chemists treat retention time as merely annotative rather than fully analytic!
I have two groups of data, Control and Test, which cluster nicely. Why is the length of the AND(Control Samples) greater than the length of the NODE(All Control Samples) in my cluster? Aren't they supposed to report the same numbers?
Not necessarily. The dynamic binning algorithm with all samples -- Control and Test -- will create a slightly different environment for comparison than if you only use the Control samples or Test samples separately. Some overlapping test or control measurements may cause the unification of markers in the dynamic binning process, or visa versa, thereby reducing the number of unique markers identified. That is, the logical AND() applied to the Control samples will report a number equal to or slightly higher than the number of markers in the cluster node representing the Control samples. This is the correct (and predicted) behavior.