Hi,
I am going to design a RNA-seq plant-insect interaction. I am not sure about how I should have biological and technical reps. If I collect the samples from 3 plants that planted in the same pot, can I consider each of them as technical reps and the whole pot as one biological rep and then the same for 2 other pots? If this is right, I would have 3 biological reps with 3 technical reps.
If I collect the samples from 3 plants that planted in the same pot,
can I consider each of them as technical reps and the whole pot as one
biological rep and then the same for 2 other pots?
That sounds reasonable. However, we don't know how biologically variable your plants are within the same pot. Commonly you speak of technical replicates if your take one sample, split it in three and generate three libraries. Biological replicates would be taking a sample from each plant and create a library for each.
Since the technical reproducibility of RNA-seq is excellent it's considered not necessary to take technical replicates, but more important to have more biological replicates to judge the natural biological variation within your groups.
This wasn't clear from your question, what are you going to compare after the RNA-seq? What is the aim of this project? Because that obviously also influences your design.
Thank you for your comment, I have two plant genotypes, one is resistant to insect and the other one is susceptible, I want to compare the gene expression of those two plants. The plants in the pots should be uniform because they are hybrid seed. Do you still think that is better to not have technical reps? The things that people usually do is to get the RNA from 3-5 technical replicates and mix them to make one library.
The things that people usually do is to get the RNA from 3-5 technical
replicates
That's for sure something you can do indeed, but it's not necessary to create identical libraries as a technical replicate.
But make sure you have enough samples in your analysis (=biological replicates). The recommended minimal is 3 samples per group, so in your case, that would be 3 resistant and 3 susceptible samples. But if you can afford it, 5 in each group is better.
We regularly get people here (on biostars) who have to "rescue" a project in which only one sample per group was sequenced (1 resistant, 1 susceptible). It's impossible to do a correct statistical analysis on that, so plan before you sequence.
Thank you for your comment, I have two plant genotypes, one is resistant to insect and the other one is susceptible, I want to compare the gene expression of those two plants. The plants in the pots should be uniform because they are hybrid seed. Do you still think that is better to not have technical reps? The things that people usually do is to get the RNA from 3-5 technical replicates and mix them to make one library.
Thanks!
That's for sure something you can do indeed, but it's not necessary to create identical libraries as a technical replicate.
But make sure you have enough samples in your analysis (=biological replicates). The recommended minimal is 3 samples per group, so in your case, that would be 3 resistant and 3 susceptible samples. But if you can afford it, 5 in each group is better.
We regularly get people here (on biostars) who have to "rescue" a project in which only one sample per group was sequenced (1 resistant, 1 susceptible). It's impossible to do a correct statistical analysis on that, so plan before you sequence.