Is the method capable of the purpose?
Is the number of replicates proper?
Is risk hedge embedded in the plan?
If you don't have any limitations in budget, time, and human resources, you doubtlessly follow statistical theory. However, you have constraints in the real world. The budget determines the maximum number of samples, which is inarguably too few, especially in microarray or RNA-Seq experiments, against the requirement from theory. If so, blindly following the principle is helpless. We dare to say from our experience on data analysis that assigning experimental parameters based on biological speculation about the background mechanism is far more crucial. Also, you have to take into consideration what the referee will ask on the submission of the paper. Subio's extensive analysis experience can help you get an adequate balance of those different viewpoints.
Inexperienced users tend to expect too much on the technology. Keep in mind that all information you heard at a conference, from fellow researchers, or makers is always biased. You shouldn't believe it as it is. Careful judgment after assessment if the technology is really capable as they say or not saves a great deal of time and money.
You know that the demo data provided by makers are usually too-good comparing to the real study data. So it is a good idea to examine the real data sets in public databases. They don't need to be of the same organism, organs or cells, diseases, etc. as your study, because your purpose is to somewhat grasp the ability, limits, and problems of the technology.
We offer the Data Analysis Service for assessment. Knowing how to analyze the data, and what output you will get gives vivid and tangible hints for your experimental planning. Even if you spent some time and money on assessment, it worth it because you can avoid far more massive loss.
When you are planning, you often ignore the fact that the experiment can sometimes fail. Yes, failure is a part of the research. We understand you want to remove all risks, though, it is impossible. So we recommend you think from two points. (1) What kind of failures are likely to happen. And (2) What type of failure is disastrous. It makes you reasonably select risks to be taken care of.
Everybody can easily imagine the second point. Contrary, the first point is difficult if you don't have a lot of experience. Moreover, failures that are likely to occur are different between experiments that you do and that you outsource. The reason you order this service is to let us complement your lack of experience with ours.
Embedding risk hedge methods in the experimental plan is not free because you have to assign some samples for this purpose. So you have to consider the cost is reasonable or not. On the other hand, there are free tricks to reduce risks, like sample labeling or how to order the outsource. Combining various means, you brush up the plan to be robust, and negative-effect-limited in case something happens.
We provide a consultant service on experimental designing, planning, and outsourcing through a web meeting and follow-up emails. Please order one Small DataSet of Data Analysis Service for it.