Run the analysis on human well-annotated reference data

Run the analysis on human well-annotated reference data to see if the distribution around the UTR end persists. This may shed some light if the shifts we are seeing are due to the annotation errors or the methodology.