Hierarchical Boosting

We have implemented a machine-learning algorithm to combine the results of multiple tests for detecting positive selection. We call this method ‘Hierarchical Boosting’, and you can find a R package implementing it in here. The details of the implementation are currently undergoing revision.

The scores are available as UCSC Hub Tracks at this address (or click the image below), or here as raw data. For a quick tutorial on how to interpret our ‘Hierarchical Boosting’ results, please go here.