Log-lin vs. lin-lin plot of OD over time

[famosab]

When extracting growth rates from growth curves in a conventional way, the log phase will be determined by looking at the curve. Thus the scale of the y-axis matters.

[famosab]

From Laura: I don't think there's a real standard method in wet lab experiment. We usually use a way that highlights the best our results of interest. CFU concentrations is expressed in powers of ten and then usually showed on log-lin plots. However it's still possible to represent it on lin-lin plots if you expect small variations rather that log variations. OD measurements (like what you did) are rather plotted on lin-lin scales since OD is not expression in powers of ten. Since you estimate your growth rates from OD measurements, I guess you should rather use lin-lin scales. But in theory, the log transformation should not impact too much the conclusions you draw from your curves, no? (Value will be different but not your trends between your conditions ?).

[draeger]

This seems to be a question about displaying the curve, or are you asking about calculation? If it is only about display: Could it be possible to let the users decice via a switch?

[famosab]

Its mainly a question of comparison to the conventional extracting method. The time points in which we see the exponential growth phase differ by display.

[famosab]

From a discussion with Simon and Laura I learned that a linear scale makes the most sense and that sometimes the OD values might be plotted on a log2 scale to easily access the doublings. I will compare those two shortly but most likely stick to lin-lin plots.

[famosab]

ToDos:

• [x] compare lin-lin and log2-lin plots
• [x] write short paragraph on OD and visualization for thesis

[famosab]

Everything has been discussed in the thesis and decided for the manuscript - closing for now.