Input:
Two counts (integers, ≥0) of the same item/event occurring in two independent sampling experiments.
Computation:
The ACD cumulative probability of observing these two counts as different as they are, if their relative frequencies are the same in the two samples (null hypothesis).
Output:
The probability value (and its logarithm) of such fluctuation. Usable as a conservative estimate of statistical significance (provided an eventually needed correction for multiple testing is applied).
1) One table of N x M counts (integers, ≥0) where N (lines) contain the numbers of occurrence of N independent rare events as drawn from M sampling experiments.
2) Two identifiers indicating which pairwise sample comparison to make.
The ACD cumulative probability of observing each pair of cognate counts as different as they are, if their relative frequencies are the same in the two samples (null hypothesis).
Ranked by increasing values (decreasing significance), the pairwise probabilities values of the fluctuations observed for each event, the corresponding distance (-ln(p)), and the count-wise normalized distance.
One table of N x M counts (integers, ≥0) where N (lines) contain the numbers of occurrence of N independent rare events as drawn from M sampling experiments.
Heatmap and matrix of MxM normalized pairwise distances between M samples.
1) An interactive heatmap of the pairwise distances between samples.
2) A table of the MxM distances (file) to be used for further analyses (clustering, tree building, etc).
