For metabolimics analysis, sample repeat is necessary, but how to know how ,any samples do I need is enough for analysis? For example, microorganism, plant or tissue?
tissue sample repeat is important, but the most important thing first, is to ensure that you neutralise any potential analytical biais. In other words, you want to make sure that the variation you observe is linked to biological variation, and not analytical biais during the sample manipulation/mode of preparation.
Once this is ensured, then it depends on the kind of analysis. When you want to do metabolomic analysis, the limiting factor is your data treatment capacity. If you do 50 samples in 6 groups for example, you will need extremely efficient super-calculators to do the job. I would recommand that you start by having a look in the literature in your domain of application to see what other scientists have done. This should give you an indication, since, depending on the domain, some may say a number which will not be seen the same in an other domain.