Slorth help

Where does this data come from?

The Slorth database is primarily designed to make predictions from the SLant classifier publically available. Data from BioGRID, the source of training data for SLant, is also inlcuded.

How are the scores calculated?

Interactions sourced from BioGRID are scored based directly on the number of references available for that interaction. Links to references are available on each interaction page.

Interactions predicted via SLant are scored based on classification confidence ranging from 0.50 which means the interaction was predicted with a very low confidence to 1 which means the interaction was classified with a high confidence.|


For yeast interactions where BioGRID cite less than 3 references the SLant confidence score has been added to the score based on number of references.

How is interaction quality defined?

Experimentally validated interactions include only those featured in BioGRID. High quality predictions include both experimentally validated interactions along with interactions predicted via SLant with a confidence score of above 0.75.