Inspect data for harvest time

In the last step you determined when to conclude the experiment. Now is time to inspect the data and identify a suitable harvest time. Here, we use onset of stationary stage as the harvest point. This is the point in time where cell concentration is no longer changing.

Step by step

1

Inspect the growth curve in Access

Pull the slider in Access to see the growth curve.

2

Find measurements with similar, high arithmetic means

Move the mouse cursor over the plot. A tooltip appears with the time and concentration numbers. Identify data points with similar, high arithmetic means. In the S. epidermidis example, this would be 11.8 and 14.9 hours.

3

Identify the max cell concentration

Find the data point that hits the highest cell concentration. In the S. epidermidis example, this would be 11.8 hours.

4

Identify harvest time

Read the corresponding incubation time on the x-axis. This is the harvest time.

You can also find the incubation time by moving the mouse cursor over the plot. A tooltip appears with incubation time (or calcHPI) as the first value.

Pick the right data point for onset of stationary phase

You have put in hard work to gather a strong dataset. It is critical that you choose the right harvest point. Otherwise your future production runs may get suboptimal cell titers.

Often it is easy to choose the best harvest point. But sometimes the answer is less obvious. In the below illustrations we provide examples of how to find the best harvest point (green) for The easy scenario and The tricky scenario. Each data point represents arithmetic mean. The y-axis is logarithmic. The error bars represent standard deviation. The transparent green boxes represent the error bands for the best harvest time data point. Contact SBT for assistance if your growth curve data do not match any of the provided examples and if you are unsure if which data point to use as best harvest time.

The easy scenario

In the below illustration, the data point with highest arithmetic mean has overlapping error bars with the subsequent datapoints. In this case, highest arithmetic mean wins and you should choose the green data point to determine harvest time.

The S. epidermidis example used throughout the MPD-1 workflow is nice and simple. The first data point in stationary phase has higher arithmetic mean than the other data points in stationary phase. This makes it easy to decide that 11.8 hours is the right harvest point for max cell titer.

The tricky scenario

The S. epidermidis growth curve example has a nice, low standard deviation. In practice the error bars are barely visible in the Access plot. But sometimes the error bars are pronounced.

The below illustration has considerable error bars. In this case highest arithmetic mean (red) does not win. The green data point represents the best harvest time. The subsequent data point has a higher arithmetic mean, but the error bars are overlapping with that of the green data point. The higher arithmetic mean for the subsequent data point may simply be caused be noise, i.e. inherent variability in sample workup and measurements.

Summary

You have now used a simplistic approach to determine the best harvest time. Next step is Summary.