Discretized probability distribution of the log-odds score under the null (background) model. Built once per motif, then queried for each observed score to obtain a p-value.
ScoreDistribution
01 Syntax
02 Methods
03 Properties
04 Members
Double)Map a score to its bin index (clamped to valid range).
Double)Upper-tail p-value: P(score >= threshold) under the null model.
Double)Lower-tail p-value: P(score <= threshold) under the null model. Useful for detecting exceptionally poor matches.
Double)Upper-tail inverse CDF (quantile). Returns the score threshold t such that P(score >= t) = tailProbability under the null distribution. Scores at or above t lie in the most extreme top tailProbability fraction of the null model and can be used as a statistically grounded minimum-score cutoff (e.g. Quantile(0.05) keeps only the top 5% of background scores).
| Name | Type | Description |
|---|---|---|
tailProbability | Double | Desired upper-tail mass in (0, 1). |
Inclusive lower bound of the binned score range.
Inclusive upper bound of the binned score range.
Width of each bin.
Probability mass in each bin; Bins[i] covers [MinScore + iBinWidth, MinScore + (i+1)BinWidth).