Custom Analysis part 3 - veeninglab/BactMAP GitHub Wiki

To quantify the movement of DnaX in the cells, I need to filter out the cells which undergo a full division, or in other words: remove the cells that don’t grow, only cover a part of a division, or are segmented wrongly. I start with the function percDivision() which filters the cells by growth speed and classifies whether a cell probably underwent a full division.

SuperSegger is a really good program for detecting division cycles, tracking genealogy and more, check their web page for more information. For the sake of testing BactMAP’s capabilities, I’ll use the function percDivision() to filter out full divisions and classify them.

perc_custom <- perc_Division(cells_custom)

perc_Division() returns a list of four elements of which the most important is a dataframe called $timelapse, which is the same as the input dataframe cells_custom, but gained a few variables which are explained here. The other three variables mean_by_percentage, plot_growth and plot_avgrowth are summarizing the results.

Since perc_custom$timelapse is a copy of cells_custom with extra variables, I’ll overwrite cells_custom:

cells_custom <- perc_custom$timelapse

Have a look at this dataframe using View() and summary().

To filter out non-growing cells and cells which didn’t undergo a full division, I need to look at the variables fulldivision and growth. Cells where fulldivision==0 did not undergo a full division, cells where growth=="none" had a linear growth coefficient (coeff) negative or close to zero. Let’s delete these cells by subsetting:

cells_custom <- cells_custom[cells_custom$fulldivision!=0 & cells_custom$growth!="none",]

Now I have a dataset which is a nice starting-point for analysis.