search

zunderlab / FLOWMAP

Other
23 stars 8 forks source link

Want to do a trajectory on cell lines, no clustering? #3

Open Stein-ErikG opened 4 years ago

Stein-ErikG commented 4 years ago

Hi, I would like to try and make a trajectory of cell lines (increasingly resistant) that is treated with different inhibitors at different concentration. I keep getting one big cluster per fcs file (mode="multi") and then a bunch og other clusters using (100 clusters) with only one cell in each cluster. Could i do a trajectory here with no clustering? Does that make sense? size.pdf

smgoggin10 commented 4 years ago

Hi, would you mind providing a bit more info about your dataset so we can make sure you are using the correct mode? To your specific question though, you do not have to use clustering -- you can do this by setting cluster.numbers and subsamples to the same value.

Stein-ErikG commented 4 years ago

Hi, I have a cell line that is either parental or increasingly resistant to an inhibitor (the parental cell line is very sensitive to the inhibitor). I have treated the four different cell lines with increasing doses of the inhibitor and in addition increasing doses of four other similar inhibitor. These inhibitors make a solid impact on the panel of signalling targets om measuring. I am interested to know if maybe trajectory mapping trough Flow-map could visualise the similarities/differences between the inhibitors and between the parental/resistant cell lines on changes in signalling. The input CSV i have made has times as the inhibitors and condition as the cell lines.

Subsamples to the same number as cluster number works great! size.pdf