Choosing qPCR reference genes and using several
A reference gene is the ruler that corrects for differences in RNA input and reverse-transcription efficiency. If the ruler itself stretches with your treatment, every target result is off.
Select several reference genes and they are combined by geometric mean automatically — and written into the Methods text.
Open the qPCR tool →1. What makes a good reference gene
- Expression does not change with your conditions (treatment, time, cell type) — the most important point.
- An expression level similar to your targets helps; very abundant transcripts such as 18S rRNA (Ct around 10) make differences hard to see.
- Good, specific primers (single melt-curve peak).
Common candidates: GAPDH, ACTB (β-actin), B2M, RPLP0, TBP, HPRT1, PPIA, YWHAZ, UBC.
2. Popular genes can change too
GAPDH is a glycolytic enzyme and can shift with hypoxia, metabolic changes or some drugs. ACTB can respond to cytoskeletal changes (differentiation, migration, morphology). Check that the gene is stable in your conditions, not just that everyone uses it.
3. Simple stability checks
- Compare raw Ct: with equal RNA input, the candidate’s Ct should be almost the same across groups; a group-mean difference within about 0.5 Ct is a practical rule of thumb.
- Dedicated algorithms: geNorm (M value), NormFinder and BestKeeper rank candidates and pick the most stable combination.
- For a new model or treatment, compare 3–5 candidates first.
4. Why and how to use several
The MIQE guidelines recommend two or more validated reference genes so that one gene’s random variation cannot drive the result.
Combine them with the geometric mean, not the arithmetic mean. Because Ct is already on a log scale, this is the same as averaging the reference Ct values before computing ΔCt (when efficiencies are equal).
Select several reference genes and they are combined by geometric mean automatically — and written into the Methods text.
Open the qPCR tool →