Hello, I have finished analyzing transcriptome level data using the TopHat>Cufflinks>Cuffmerge>Cuffdiff flowthrough. Looking at the assembled transcripts in a genome browser (IGV, IGB or Savant), I consistently notice differences in the start and end sites of the assembled transcript versus the annotated reference one. Can somebody tell me the meaning of this? Is this due too low scoring at the end of the transcript? Thanks. GP.