Cuffdiff Runtime and Results Consistency

Hello. I ran Cuffdiff on some RNAseq data last weekend, and it completed in about 48 hours. However, I discovered that I didn’t select the option to generate the SQLite file for use in CummeRbund. So, I re-ran the same analysis, but it’s been running for four days now. I have two questions:

  1. Should it take this long to run?
  2. How consistent are Cuffdiff’s outputs between runs? Will genes that came up as differentially expressed from the first time I ran Cuffdiff have the same results the second time around (ie does Cuffdiff use any random error to generate its results)? Or do I need to wait for it to finish re-running to start my follow-up analysis?

Thank you!

Hi @elizamccoll
maybe refresh Galaxy page by clicking at “Galaxy” banner at the top left corner. Is the job still running?
Four days: do you mean run time or both queue time and run time?
The run time depends on resources allocation for a job.
In my experience Cuffdiff results are consistent, as long as you use the same version of the software.
Kind regards,
Igor