Skip to main content

RNA Purification: Common Pitfalls and Fixes

Team Dynamic Matrices | 15 May 2026

We have looked at a lot of failed RNA preps, ours and customers'. The same ten problems account for almost all of them. Each one below has a sign you can see in your QC data, a cause, and a fix. Keep this next to the bench.

1. Low yield from a sample that should have plenty

Sign: Fluorometric concentration far below expectation; A260 looks fine or is inflated.
Cause: Incomplete lysis (tissue not homogenised, cells clumped), overloaded column, or RNA lost in the interphase during phenol extraction.
Fix: Homogenise harder or longer; stay within the column's binding capacity (split the sample); after phase separation take only the upper aqueous phase and leave a margin above the interphase. For fibrous or fatty tissue, add a proteinase K step.

2. RNA degraded despite good technique

Sign: RIN under 7, smear on the Bioanalyzer, 28S:18S under 1.5.
Cause: RNase from the sample itself (pancreas, spleen, skin are RNase-rich), delayed stabilisation, or contaminated surfaces and tips.
Fix: Flash-freeze or put tissue into a stabilising reagent within minutes of collection. Use a chaotropic lysis buffer at once. Dedicate an RNA bench, wipe surfaces with an RNase-decontaminant, and use filter tips. Keep samples on ice only after lysis; before lysis, cold does not stop endogenous RNase fast enough.

3. Genomic DNA contamination

Sign: A260/A280 looks normal; no-RT qPCR control gives a signal; RNA-seq shows intronic and intergenic reads.
Cause: gDNA co-purifies on silica and in phenol methods. This is not a failure; it is the default.
Fix: On-column DNase, or in-solution DNase followed by cleanup. Confirm by no-RT qPCR against a genomic target, every time. Design RT-qPCR primers across exon junctions.

4. Phenol or guanidine carryover

Sign: A260/A230 under 1.8; RT and downstream enzymes inhibited; low library yield despite adequate RNA.
Cause: Insufficient washing, or wash buffer not fully drained before elution.
Fix: Add a wash; spin the empty column for a full minute before elution; for phenol methods, do a second chloroform extraction. If A260/A230 is already low, re-purify on a column.

5. Salt in the eluate

Sign: A260/A230 low; ethanol precipitation pellet is glassy and large; enzymes inhibited.
Cause: Wash ethanol not removed, or elution buffer contaminated.
Fix: Ensure the final wash is drained and the membrane is dry. Elute in nuclease-free water or low-TE.

6. Concentration overestimated

Sign: Absorbance says 200 ng/µL; Qubit says 60 ng/µL; libraries under-perform.
Cause: UV absorbance counts free nucleotides, gDNA, salt and phenol as RNA.
Fix: Quantify by fluorometry (RiboGreen or Qubit RNA HS). Use absorbance only for purity ratios.

7. Small RNAs lost

Sign: miRNA qPCR fails; small RNA-seq libraries empty; Bioanalyzer shows no small RNA peak.
Cause: Standard silica columns and many precipitation protocols lose RNA under 200 nucleotides.
Fix: Use a kit that specifies small RNA retention (usually a higher ethanol concentration at binding) or a phenol method with a high-salt precipitation and a glycogen carrier.

8. Wrong QC metric for the sample

Sign: RIN reported as 2 for an FFPE sample, or for an EV prep, or for bacterial RNA.
Cause: RIN assumes intact eukaryotic 18S and 28S peaks.
Fix: Use DV200 for FFPE, small RNA electrophoresis for EVs, and a bacterial-specific algorithm for prokaryotes. Our QC checkpoints post covers this in detail.

9. rRNA dominates the library

Sign: 30 to 60 percent of reads map to rRNA after "depletion".
Cause: Depletion chemistry mismatched to the sample: probe panel does not cover the organism's rRNA variants, or RNase H digestion incomplete on fragmented input, or input outside the validated range.
Fix: Check the depletion method against your sample type and input. For degraded, low-input or non-model samples, use a method that captures rRNA by sequence without depending on enzymatic digestion. Confirm depletion by qPCR before sequencing.

10. Variability between replicates that were "handled identically"

Sign: Technical replicates diverge in yield or integrity with no obvious cause.
Cause: Time from sample disruption to lysis varied; wash timing varied; a new kit lot.
Fix: Record time-to-lysis and keep it constant. Time the washes. Log kit lot numbers in the data table so they can be a covariate. See our post on why RNA workflows fail reproducibility.

The one-page checklist

  • Stabilise within minutes of collection
  • Lyse in chaotrope immediately; do not rely on ice
  • Stay under column capacity
  • DNase treat, then confirm by no-RT qPCR
  • Drain washes; spin the dry column
  • Quantify by fluorometry, check purity by absorbance
  • Choose the QC metric for the sample type
  • Match depletion chemistry to input and sample; confirm before sequencing
  • Log time-to-lysis and lot numbers

References

  1. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162, 156–159 (1987).
  2. Schroeder A et al. The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Mol Biol 7, 3 (2006).
  3. Fleige S, Pfaffl MW. RNA integrity and the effect on the real-time qRT-PCR performance. Mol Aspects Med 27, 126–139 (2006).
  4. Bustin SA et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55, 611–622 (2009).
  5. Adiconis X et al. Comparative analysis of RNA sequencing methods for degraded or low-input samples. Nat Methods 10, 623–629 (2013).

FAQs

A260/A280 (target ~2.0) flags protein and phenol; A260/A230 (target over 1.8) flags guanidine, salt and phenol. Low A260/A230 is the more common cause of enzyme inhibition.

For RT-qPCR and RNA-seq, yes. For northern blots and some hybridisation assays, gDNA is tolerable. When in doubt, treat.

Standard total RNA library kits want 100 ng to 1 µg. Low-input kits go to 1 to 10 ng. Below that you are in single-cell territory and the chemistry changes.

Flash-freeze tissue or put it into a stabilising reagent within minutes of collection, lyse straight into a chaotropic buffer, use a dedicated RNase-free bench and filter tips, and keep samples on ice only after lysis. Then check integrity by capillary electrophoresis (RIN, or DV200 for FFPE) before library prep rather than after a failure.

Use an on-column DNase step, or an in-solution DNase followed by cleanup, and confirm removal with a no-RT qPCR against a genomic target every time. A normal A260/A280 does not rule out genomic DNA.

Explore riboLASSO →

Related Reading