What are the best ways to design a discovery and replication association study in GWAS or EWAS?
Should the total samples available within your funding budget be split into two for discovery and replication?
If so, should they be measured using different technologies? OR could you use the same chip/array. If you could, would you run them in different batches or all together.
An finally, how large should the replication study be relative to your discovery population?
Thanks in advance for your help everyone.
I am not an expert in the field. I am learning principles of Epi in genomics research and I can provide some input based on my current knowledge.
A good study will include a large number of samples in discovery phase and include an appropriate number of additional samples (probably new samples) in replication phase. The preferred method is Sanger's sequencing for replication phase.
These are not strict rules. I have seen people including additional samples along with samples used in GWAS for replication studies.
Power calculations will tell you the appropriate number of samples to be included in discovery and replication phase.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics