Concept and Background

Arboretum was conceived to address the need for a faster alternative for the classic GENIE3 implementation for inferring gene regulatory networks from high-throughput gene expression profiles.

In summary, GENIE3 performs a number of independent learning tasks. This inference “architecture” suggests two approaches for speeding up the algorithm:

  1. Speeding up the individual learning tasks.
  2. Specifying the task coordination logic so that the tasks can be executed in parallel on distributed hardware.