What is the minimum number for biological and technical replicates for a reliable results of RT-qPCR? what if I collect/gather all the biological samples from different replicates of the same groups in only one sample?
The concept behind biological and technical replicates is that we use technical replicates from the same sample to ensure the consistency of the measurement tool and technique as you use portions from the same sample and expose them to the same methodology and tool to measure the random noise that can be due to the protocol or the instrument itself. On the other hand, we use biological replicates to ensure the reproducibility of the results as you use different samples "biologically close/similar" to assess the random biological variations.
The more the number of replicates you use the stronger powerful statistics you get and the less error you may encounter.
I would recommend triplicates as the minimum but due to limited samplings and amounts of each sample, usually duplicates work well too.
In our recent NASA-Space flight experiment (small amount for recovered samples), we decided to use 2 Biological replicates + 2 technical replicates for each biological ones (in total 4 replicates for each sample). These samples was used for qPCR, and Microarray analysis ( CATMAv6.2 array).
Regarding to collect all the biological replicates in only one sample, You may Error of your experiment, but what about the qPCR reaction variation? Of course, a technical sample should solve this part!
Normally, we use a pool of all the collected samples, replicates in one sample as an internal biological (Genomics) control!
Do you want to simply mix together samples from different biological replicates to save space in PCR plate? Well, in this case this sample would contain the average cDNA concentration of your target for the group, so you will see the average result anyway. But you will lose all information about variations within the same group, so your result would be of no use for any statistical analysis. I strongly dissuade you from doing this. For statistics, 8 and more biological replicates is usually considered good, but if they are hard to prepare, you will need a minimum of 3 in each group.
As for technical replicates, usually you will need 3 for each biological sample (also for positive and negative controls), and in a pinch it may be reduced to 2. You basically only need them to make sure that your reaction is reproducible. If there were some problems in particular wells you may exclude them from analysis, and it's easier to decide which to exclude if you have one outlier out of three replicas than if you have 2 different numbers.
- Badges
- Science method