Very low assigned reads in featureCounts

Hi @Danni

Let’s combine all three questions together.

  1. Low annotation assignment rates Very low assigned reads in featureCounts
  2. No/low differential expression Differential gene expression analysis with only 2 biological replicates
  3. Reference genome + reference annotation Rat reference genome selection

Possible reasons for these results.

  1. Reference genome and reference annotation are not based on the same exact genome assembly. Or, one or both have a formatting problem. Much Q&A about how to confirm and resolve this problem are at this forum. That includes how to source the data. I added some tags to your post to help find these, and many include GTN FAQs.
  2. The reads are lower quality, or possibly did not have QA steps run, or are actually not from the same species as the reference data. This topic also has much Q&A, including for this specific tool, and I’ve added tags.
  3. If all of the above checks out, and you are still not getting any differential expression reported by one tool, you can test out other DE tools to compare.
  4. In the end, it is possible there is no detectable expression difference between the sample groups using these methods. The reasons could be related to the number of samples (3 per group is the minimum suggested by the tool authors). Or, problems could be upstream: library preparation problems, sample sequencing problems, or not biologically relevant “different” sample groups in terms of transcriptome expression levels.

Please give those resources a review.

  1. Correct or fix reference mismatch or content or format issues.
  2. Perform QA on the reads, and decide if those have acceptable quality.
  3. Consider viewing all files in a browser like UCSC or IGV. Drill down into a few known genes of interest. If you do this at UCSC, turn on more tracks for context. This can reveal even complex problems very quickly.
  4. Run through a few tutorials that demonstrate the methods/tools for this analysis domain.