Figure 6.

Model performance on proteins identified by affinity proteomics. (A) Proteins detected solely by PEA and not mass spectrometry have a lower average brain abundance according to PaxDB. (B) The classification models perform well on the CSF proteins identified by affinity proteomics, identifying the majority of them. Importantly, the models are able to correctly predict low abundance proteins that are potentially only identifiable by targeted approaches. PEA – proximity extension assay; PPM – parts per million.