Does stemming affect parser output?


I need to preform both stemming and parsing on a block of text. Stemming would be easier to apply before parsing by adding it to the block of filters used to prepare the input for parsing and POS tagging. However if doing this would decrease the accuracy of the parser then I would want the stemming to happen after the fact.
If stemming does need to be done after parsing, what would be the best way to convert the parse tree into a token stream and back again?