Patrice Salomé of Plant Editors lays out the difference between biological and technical replicates with an easy-to-digest analogy.

You just found a new cookie recipe you want to try out. Before serving an untested baked good to your friends and family, you decide to make one batch to make sure they are as good as the recipe author claims them to be.

To your dismay and horror, the first cookie of your first test batch tastes terrible! 

What would you do next? Try your luck with another cookie from the same batch? What makes you think it will be any better than the first one you tried? Make another batch with the same exact components you have in your fridge and pantry, in case you got the amounts and proportions wrong? Go to the store and buy fresh eggs, milk, sugar, and other ingredients? Surely one of the ingredients must have turned bad and spoiled the entire batch.

Now let’s rename cookies and batches in terms of technical and biological replicates. The first batch was one biological replicate; each cookie from that batch is a technical replicate, as they represent non-independent samples of the batch. No matter how many cookies from that batch you try, they are likely to all taste the same. If you were to make another batch of cookies with completely different ingredients, this second batch would be a second biological replicate that is completely independent from the first batch. The situation becomes murkier if you make another batch of cookies using the same ingredients as the first batch; the two batches are clearly not independent, and are likely to result in the same outcome (ie, a bad cookie). This is an example of pseudoreplication.

When designing experiments, we strongly advise to follow a 3 x 3 (or even 3 x 3 x 3) rule: perform the experiment completely independently three times (these would be the biological replicates [or independent batches of cookies), with each experiment comprising three samples (the technical replicates [or three cookies from one batch]). The addition “x 3” is a replication of the assay (like taking three bites out of one cookie).

For example, let’s say you want to test the expression of your favorite gene in wild type and a mutant. For each biological replicate, you collect three samples (aka technical replicates) for the wild type and three samples for the mutant (each sample being different plants or seedlings grown on different plates). You extract RNA, reverse-transcribe it, and then quantify expression levels by quantitative PCR, using the above cDNA as template. Because PCR is likely the only technique that does not follow the adage “repeating something and expecting a different outcome is a sign of insanity”, you should very much set up more than one reaction per sample per gene; a good number of such technical replication is three, hence the third “3”.

So in short, when planning your experiments, think in terms of cookies if you are unsure of the level of replication!

Replications done and ready to publish? Contact us at planteditors.com/sales for scientific editing and image analysis!