Inference procedures
====================

Quantifying within-sample heterogeneity
---------------------------------------------

Memento uses bootstrap to quantify within-sample heterogeneity of gene expression. You can specify whether to use the raw values to calculate a P-value or fit a Normal distribution to the sampling distribution. We have found that this works particularly well for the mean.

Quantifying between-sample heterogeneity
---------------------------------------------

**Hierarchical bootstrap**

Under construction

**Quasi-maximum likelihood**

Under construction