SPRAS Benchmarking

SPRAS is a utility software designed for performing signaling pathway reconstruction with various algorithms. SPRAS’s documentation has more information about its inner workings software.

This benchmarking repository (see the GitHub) is meant to display the performance of all of the algorithms currently supported by SPRAS on signaling pathways and diseases (to test out DMMMs), comparing their reconstructions with manually curated golden datasets, or synthetically provided datasets from various databases.

All information provided is orchestrated through our GitHub Actions pipeline, and heavy processing is soon to be moved to HTCondor.

Format

Each run’s slug has the type, the dataset, the algorithm, and the paramaters hash.

There are also pages related to different categories of these runs:

The type classifies a dataset and the algorithms it runs on. For example, PRA datasets run on all algorithms, while disease module datasets only run on DMMM algorithms.

Runs (11)