identifying positive events based on intensity data using R

Question

I would like to ask the community for any ideas of packages in R that I could use to do automatic gating and to quantify the number of events I got outside this gate. I have many different datasets with intensity values as the pictures attached. In this datasets there is always a baseline intensity around 1000 (as showed in the pictures attached) but it varies slightly in every file; i.e in the first pic this gate would be from 0 to 1200 approx, while in the second pic would be 0-1800 or 0-2000. Counting the vents above this thresholds for picture 1 and 2 we can see pic1 has less positive cells than picture 2.

I would like to apply any model/algorithm to identify this baseline individually for each dataset and to count all the events outside this baseline (positive cells). Is this possible? I tried mixed model before with two Gaussians (Mclust package) but I think is not the best approach as the baseline is a Normal Gaussian, but the positive cells is a skewed population, and I was trying Cytometree package using the binary tree option but it doesn't identify 2 populations, only a single population, don't know why, maybe the 2 populations are not enough separated for this. Another option could be treating these positive cells with high intensity as outliers? and try to score/quantify outliers?

Any ideas?

Thanks in advance

question from:https://stackoverflow.com/questions/65938313/identifying-positive-events-based-on-intensity-data-using-r

Answer 1

Waitting for answers