Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: Integration of probabilistic functional networks without an external Gold Standard

Fig. 2

The BioGRID Saccharomyces cerevisiae data increase. A High-throughput (HTP) interaction data has increased since our earlier study [8] (yellow box; dashed line), with two large increases in 2010 and 2016 (grey boxes). By contrast, low-throughput (LTP) data has increased to a lesser extent, with the overlap between HTP and LTP also increasing slightly. B The fluctuation in representation of unique proteins that was observed in our earlier study [8] (yellow box; dashed line) is not evident in the following years. Unique protein number has increased steadily for both HTP and LTP data and the large increases in interaction data (grey boxes) only affect this increase slightly. Version 186 LTP data contains ~1/3 of the total unique proteins in BioGRID, with LTP data almost completely overlapping with the HTP dataset. C The size and overlap of interactions in V186 HTP datasets and LTP Gold Standard data as the Gold Standard size threshold is increased. D The size and overlap of proteins in V186 HTP datasets and LTP Gold Standard data as the Gold Standard size threshold is increased. An LTP dataset size threshold of 100 interactions (dashed line) was chosen for subsequent analyses to maximise both the number of the LTP Gold Standard’s unique proteins and unique interactions, while minimising its overall size

Back to article page