Anyone performing industrial functions, including antibiotic production, beer brewing, production of high-value industrial chemicals, development of analytical devices for screening infections, and assessing contamination of fresh vegetables, depend upon accurately characterizing the microorganisms employed for those functions. A key element of that characterization is measurement of the growth rates of those organisms.
In recent years the utilization of microplate readers has permitted high-throughput growth measurements at unprecedented rates, but the manual analysis of the resulting data provided a serious bottleneck in conversion of the data into useful growth rate determinations.
Analyzing Microbial Growth
Growth of microbial cultures is determined by measuring the population density of a culture at a series of time points by determining the turbidity of the culture at each time point. Typically cultures were grown in flasks or tubes, and at the required intervals a sample was removed and the turbidity measured in a spectrophotometer cell. About 18 cultures was the most that a technician could monitor at one time, and doing so required the technician’s full time effort of the course an 8 -10 hour work day. Today up to 384 cultures can be grown simultaneously in a microtiter plate, and a microtiter plate reader can read the turbidity of each culture at specified intervals automatically without any attention from a technician, and store those readings for further analysis to determine the growth rates of the 384 cultures. Setting up such a growth rate experiment requires only about 3 hours.
Manual analysis of the resulting data, however is both tedious and error prone.
Data analysis of a single culture requires multiple steps:
- Subtract the background OD from each time point
- Take the ln of each corrected OD.
- Plot those values on a graph and decide which points fall along a straight line.
- Copy/paste those time points and ln corrected ODs into a statistics or spreadsheet program.
- Determine by linear regression the slope of that line, which is the growth rate for that culture. An experienced person could analyze the data for one culture in several minutes, so about 4–8 working days would be needed to analyze all 384 cultures. While that is considerably better than the roughly 30 working days required for data collection and analysis of 384 cultures by the old shake flask method, that is hardly consistent with the concept of “high throughput”.
Let the Programs Do the Work
We’ve developed a suite of tools including GrowthRates 6.2, that takes the output of a microtiter plate reader, does all of the calculations described above and in addition to the growth rate calculates the standard error of the growth rate and a correlation coefficient as estimates of the accuracy and of that rate confidence in the rate.
We are aware of other programs that automate the calculation of growth rates from plate reader data, but each of those requires significant programming skills in MATLAB, R, or Python. GrowthRates requires no programming skills. Plate readers output their data in at least 16 different formats (so far). All of the other programs require manually reformatting the plate reader output for input to the program. GrowthRates automatically translates those formats to a standard format.
Subscribe to receive updates and new release information.