Remove super-expressing cells from analysis?

[jstitlow]

Super-expressing cells are slowing down the analysis and I'm not confident that they are being quantified accurately.

Any ideas on how to remove these cells from the quantification within bigfish?

My thought would be to have a segment-first mode where you segmented the cells first, and then run spot detection on each cell individually. But I don't know how much this would slow things down.

The super-expressing cells could then be treated as one big cluster, and then the number of spots could be estimated based on integrated density of an average single spot.

Is this problem too obscure?

[muellerflorian]

Hi Josh,

your problem doesn't sound obscure. the proposed solution is also interesting. however, big-fish is currently designed to detect spots in the entire image and will then assign them to cells in an (optional) postprocessing step.

i think that inversing the process would require a substantial reorganization of the code, which will be difficult.

another angle would be to understand why the code is so slow (your issue #44). if we could solve this, we could remove highly expressing cells during the post-processing step.

So I would propose that we focus on this first.

[Henley13]

Hi @jstitlow ,

In addition to @muellerflorian's comment, if you do not provide a 'threshold' parameter for your spot detection, the algorithm will try to find one automatically and this process works better with multiple spots in the frame. If you run spot detection cell by cell there is a risk that the threshold automatically set won't be accurate enough (if there are not enough spots in the cell).

Otherwise, one quick solution to test would be to:

1. Segment cells.
2. Loop over cell instances.
3. for each instance, mask out every other cells in your filtered FISH channel (set pixels to 0).
4. for each instance, run spot detection.