Is it important to take the same replicates for control and treated samples?

Question

[nivya.james2016]

Hi all,

I am new to NGS analysis. I have this doubt; is it a must to take the same replicates for all the reads (control and treated) for performing the miRNA Seq analysis? (eg: is it must to take it like this:- rep6 of read 1 and rep6 of read 2?)

Kindly help me.

Thank you in advance

[h.mon]

It is impossible to know exactly what you are asking. I will make a guess, and suppose you are asking if it is necessary to have the same number of biological replicates for control and treatment.

No it is not necessary, but it is better if: 1) you meet a certain number of biological replicates, a good rule of thumb being at least five, 2) the number of biological replicates is not very different between treatment and control, 3) the number of sequencing reads per replicate reach a certain minimum, another rule of thumb here being 2million reads per sample.

[nivya.james2016]

Thank you so much.

One more query i have is; is it necessary to take same number of control and treated samples. For example: If i am taking two treated samples is it necessary to take two control samples?

[ATpoint]

It is advisable to have your groups balanced, but money, time and availability of samples often interfere with this. In any case, you should have replicates per group. If possible more than two per group. Statistical power increases with both sequencing depth and replicate number. The fewer replicates you have, the more dramatic the changes between conditions need to be in order to detect them at a given significance level.