Custom Analysis Part 4
For snap-shot data, I am used to making demographs: localization plots
based on cell length. For time-lapses, it can also be very useful to
combine the data of all cells together, but instead of cell length, I’d
rather use cell division. To be able to do this,perc_Division() takes
the cell division time of each cell as a percentage of division where
0== cell birth, and 100 equals the point where the cell produces 2 daughter cells. In this
way, it is possible to combine and plot all cells, no matter their size
or growth speed, based on their division.
To make a kymograph based on division percentage, first combine the dataset we just made,
cells_custom, with the dataset containing the TIFF image, image_custom, so the pixel values are connected to the cell information.
cells_image <- extr_OriginalCells(image_custom, cells_custom)Now, I can use bactKymo() to plot the fluorescence over division.
bactKymo() has the option percDiv.
When put to TRUE, the cell fluorescence will be binned per division
percentage.
bactKymo(cells_image$rawdata_turned, timeD=TRUE, percDiv=TRUE, sizeAV=TRUE, mag="100x_DVMolgen")
Even though I already discarded non-growing cells & cells with a
too-short division time, I decided to go through the kymographs and see
if the quality of the kymographs is high enough. I only selected on
growth using perc_Division(), but there also might be cells without
detectable fluorescence, or wrongly segmented cells, and I see no other
way to check for these than by eye, unfortunately. To do this as well,
save the PDF of seperate kymographs again:
cairo_pdf(filename="Kymos_cluster.pdf", onefile=TRUE)
bactKymo(cells_image$rawdata_turned, timeD=TRUE, sizeAV=TRUE, mag="100x_DVMolgen")
dev.off()Based on this, I selected a group of cells to discard. Some cells did
not grow, some have a very low fluorescence, and some only have
fluorescence on the edge of the cell. I saved their cell numbers in
classification_eye.csv. For the tutorial, I suggest you do your own
selection (or if you wish, skip this step below).
I remove the cells from cells_custom and cells_image:
class_eye <- read.csv("Fig5D-H/classification_eye.csv", header=TRUE)
cells_custom <- cells_custom[!cells_custom$cell%in%class_eye$cell,]
cells_image$rawdata_turned <- cells_image$rawdata_turned[!cells_image$rawdata_turned$cell%in%class_eye$cell,]| ⬅️ Custom Analysis Part 3: Filtering the Data | Custom Analysis Part 5: Preparation for Cluster Analysis ➡️ |
|---|
