Note that patterns must follow Tgrep2 and Stanford Parser/Penn Treebank syntax. Here is some example output from the Stanford Parser. Also, make sure to put "S1|ROOT <<" at the beginning of a pattern to reduce duplicate matches.

Example Pattern: S!>>NP !>>VP <NP <(VP< (/^VB/ << be|was|is|were|am|are|being|been) <(VP<VBN))