For example, we have 100 samples, among them 50 are resistant, others are susceptible. Traditionally we sequenced 100 samples one by one ,and then call SNPS and other analyses. I wonder whether is it ok to seperate the 50 resistant samples into 10 groups for example, and then sequenced the groups as a total. If so ,we just need to sequence 10 groups which is cost effective and i think this will not influence too much for subsequent analysis. Anyone have published similar ideas in paper or have discussed it?
Maybe for humans,we need more accurate and more accurate. However for practical uses, we need to banlance. Especially, for a large samples, such as 50000 or more?