In this software demonstration, we will show how formal ontologies can be used to label in-stances of neural (ERP) patterns that have been extracted from multiple datasets using a novel pipeline for pattern and metric extrac-tion (see Figure 1, next page). The entire dem-onstration will last ~15 minutes. We will begin with a 5--minute introduction to ERP data from several cross--laboratory studies of word com-prehension. This overview will motivate our demonstration by showing that ERP data are complex and heterogeneous, which explains the radical challenge of making valid compari-sons across different studies within our do-main. We will then give a 2--minute description of the pipeline for analysis, which has two main components: (1) a set of pattern extraction (signal decomposition, temporal segmentation) methods; and (2) code to extract a variety of simple metrics (e.g., min and max intensity at a particular electrode) and to express these summary features as N--triples, which are sub-sequently stored in RDF. Finally, we demon-strate how the NEMO ontology can be used to reason over these data. We highlight both ex-pected and novel findings for the test datasets and note that large--scale application of this method could lead to major breakthroughs in understanding neurological patterns that are linked to sensory, motor, and cognitive pro-cesses in neurologically healthy and brain-injured children and adults.