After analysis on the galaxy server using CuffDiff, all my data is insignificant based on the q value, even if p-value is significant for all genes showing differential expression. What could be the possible explanation for the same? And how can this data be used for further validation?