Tweet
Hello,
I've recently started working in a group that would like to start doing RNA-Seq experiments in order to compare the transcriptomes of pancreatic islets undergoing different treatments.
As I'm new to this kind of work, I was wondering if anyone could offer advice on experimental design - specifically, how do you choose the correct number of samples (technical and biological replicates) and the depth of sequencing to give statistically meaningful results? Is it necessary to carry out a 'power' analysis to answer these questions? If so, how do you do that?
I've spent some time going through the literature on this, but frankly, it's quite confusing and seems inaccessible to anyone without a background in advanced maths or statistics. (I have neither.)
However, on the basis of what I've read, I've decided to run a pilot sequencing experiment with a limited number of samples and then enter the data into the 'Scotty' web-tool as described here:
This then returns information on number of samples and read-depth required to achieve a specific statistical 'power'.
Does that sound like a reasonable course of action?
Any advice on this matter would be much appreciated. In fact, any advice, or pointers to worthwhile reading material/resources in the area of RNA-Seq difference analysis would be a huge help.
Thanks for any advice offered.
Regards,
John
I've recently started working in a group that would like to start doing RNA-Seq experiments in order to compare the transcriptomes of pancreatic islets undergoing different treatments.
As I'm new to this kind of work, I was wondering if anyone could offer advice on experimental design - specifically, how do you choose the correct number of samples (technical and biological replicates) and the depth of sequencing to give statistically meaningful results? Is it necessary to carry out a 'power' analysis to answer these questions? If so, how do you do that?
I've spent some time going through the literature on this, but frankly, it's quite confusing and seems inaccessible to anyone without a background in advanced maths or statistics. (I have neither.)
However, on the basis of what I've read, I've decided to run a pilot sequencing experiment with a limited number of samples and then enter the data into the 'Scotty' web-tool as described here:
This then returns information on number of samples and read-depth required to achieve a specific statistical 'power'.
Does that sound like a reasonable course of action?
Any advice on this matter would be much appreciated. In fact, any advice, or pointers to worthwhile reading material/resources in the area of RNA-Seq difference analysis would be a huge help.
Thanks for any advice offered.
Regards,
John