The graph specifications control what metric is used on the Y-axis, and whether smoothing will be applied.

When you click the Options disclosure triangle in the kinetics window, the options shown below appear. They modify the attributes of the current graph (or currently-selected gate.)

The ‘Statistic’ selection is the statistics that will be applied to the parameter selected for display on the Y-axis. There are two basic categories, statistics and statistics of events that exceed a threshold. The statistics that can be selected are:

- Median: the median fluorescence of the events in each time interval, either of all events or all events exceeding some threshold, annotated as (for cells > T).
- Mean: the mean fluorescence of the events in each time interval , either of all events or all events exceeding some threshold, annotated as (for cells > T).
- Geometric mean: the
*n*th root of the product of multiplying all*n*events within the time range together, either for all events or all events exceeding some threshold, annotated as (for cells > T). - Percentile: An arbitrary percentile of the events in each time interval, for either all events or all events exceeding some threshold, annotated as (for cells > T). Entering 50 in the percentile value generates the same graph as Median.
- % of cells over Threshold T: The percentage of the events that are greater than the set threshold in all ranges, for the selected Y-axis metric. The threshold is a scale value of the Y axis parameter (i.e., not the channel value, but the scaled linear fluorescence value). The threshold can either be an absolute value, or a relative value based on time slices that you have defined (see information on defining thresholds).

The two possible smoothing functions are Moving Average and Gaussian. They are similar. The moving average takes a small range of the data centered at the bin in question, calculates the average value and displays that for the bin. For the next bin, the range across which the average is calculate is shifted by a bin so that the new bin is centered. The moving average calculation moves across the data in this manner. Gaussian smoothing operates in the same manner except that more weight is placed on the bins in the center of the calculation, and less weight on values more distant from the center, with the weighting following a Gaussian distribution.

*See also:*