[Full Guide]: Molecular Biology Fails & Fixes (PCR, Cloning, and Preps)
Textbooks only cover basic issues (bad primers, wrong T°C). But they don’t talk about the realities of the bench.
Wildtype One searched every PCR, cloning, or plasmid prep failure reported by researchers.
Below is your full guide to fixing your molecular biology experiments.
Print it
Hang it in your laboratory
Share it with colleagues who might be struggling as well
Let’s turn “WTF” into “Aha!” moments.
Article Sections
I. PCR
II. Cloning
III. Minipreps
IV. Which kits to buy? (company-specific insights)
V. Data & presentation pitfalls
Part I. PCR Nightamre Failures (And Easy Fixes)
“Molecular is half science, half voodoo!”
One frustrated grad student exclaimed.
I. 1. The Vanishing Act
Scenario: Your PCR worked 200 times, then suddenly only smears—even in the no-template control (NTC)!
“I tested over 200 samples with a nice clean band until one day it suddenly stopped... a long smear, even in the NTC”.
They changed everything (water, primers, polymerases, gel, cycles) – smear still haunted them.
I. 2. Contamination Culprit
…and it’s often a simple issue.
In the previous case for example, a commenter said:
“Had something very similar... an undergrad contaminated the pipette. We switched to filter tips and the issue went away.”
Yes—a possessed pipette. DNA aerosol from previous aspirations can seed your PCR.
What you can do
Always use filter tips (and consider cleaning or dedicating a PCR-only pipette).
I. 3. Nucleases or “Cursed” Primers
If one primer set shows smears but others work fine, they might be degrading over time or forming weird artifacts.
If you resuspend primers in water, ensure it’s truly nuclease-free (commercial or freshly autoclaved).
In stubborn cases, dissolve primers in TE buffer (Tris-EDTA) to protect from degradation.
Store small aliquots to avoid repeated freeze-thaw cycles that “dephosphorylate” or otherwise damage primers.
I. 4. Tricky Template Regions
The most important targets often put up the biggest fight.
Sometimes, the genomic region itself is jinxed.
High GC content
Secondary structures
Repeats
All can stump the polymerase.
What you can do
Add DMSO or betaine to help denature stubborn secondary structures
Use touchdown PCR (gradually lowering annealing T°C) to coax out a product
If all else fails, design new primers. Make them a bit longer or in slightly different spots
One colleague on ResearchGate found that:
“Using primers ~10 bp longer worked perfectly… the one time it worked with old primers was a lucky hit.”
I. 5. Check the Usual Suspects
It sounds basic. But when PCRs go awry, check fundamentals:
Temperature drift or uneven heating can cause silent fails.
Old or imbalanced dNTPs can cause smeared, partial amplification. Ensure each dNTP ~200 µM; too high can chelate Mg²⁺ and wreck amplification.
Too much template DNA (>50 ng per 50 µL) can also inhibit PCR—dilution can be a solution.
If you use an aging machine, verify its calibration. Our troubled researcher tried three different old thermocyclers. If available, run a known-good control on the same machine (even a pre-validated PCR from a colleague).
I. 6. When the Gel is the Villain
You might have good PCR products, but the gel electrophoresis lies to you!
Symptoms: ladder looks wonky, bands “smear” or stay in wells despite successful amplification. Believe it or not, mistakes like using TAE buffer in the gel and TBE in the tank can cause bizarre patterns.
What you can do
Ensure you use the same buffer for gel prep and running
Check that the gel solidified evenly (no tilted gel tray) and the electrodes/intact wires on your apparatus
Running gels too hot and fast can melt agarose, or too slow can let DNA diffuse, both yielding smeary results
If your NTC shows a band or smear, the problem isn’t the PCR target—it’s contamination or a detection artifact. Pro tip: NanoDrop your water to check and measure DNA absorbance—Do you see a peak at 260 nm? Red flag!
I. 7. One Final Real-World Tip
If you’re tearing your hair out, get a second pair of hands/eyes.
Back to the previous PCR failure, a colleague re-ran PCR with different pipettes, water, tubes. Same smear result—ruling out a single-person error. If you have a similar case, it’s not your fault.
Another labrat’s colleague suggested replacing the DNA loading dye, and it unexpectedly fixed PCR bands “stuck” in wells.
Sometimes two heads (and fresh reagents) are better than one!
I. 8. Summary of Common PCR Gremlins and Fixes
Part II. Cloning Catastrophes: Ligation, Transformation, and Colony Woes
Did you ever cut, ligated, transformed—but instead of glorious colonies, you get nothing? Or mutants? Or just empty vectors?
Let’s troubleshoot some cloning horror stories:
II. 1. The No-Colony Curse
You get zero colonies on your plate (except maybe the pesky background on the vector-only control).
Countless students reported struggling for months on a single cloning.
They did everything by the book:
Confirmed insert and vector digests
Used fresh ligase
Ran ligation checks on a gel
Saw the expected concatamers
…and still nothing grew after the transformation
II. 2. Incompetent Cells
In one case, troubleshooting pointed to the competent cells. The student’s viability test (transforming a known plasmid) yielded <50 colonies when there should be hundreds.
Fellow researchers immediately chimed in:
“Clearly your competent cells are not so competent.”
Low-efficiency cells will kill your cloning. Our student even made fresh DH5α stocks three times, to no avail. It turned out the lab’s cell stock was likely old and tired (or repeatedly thawed).
What you can do
Keep cells ultra-cold, avoid multiple freeze-thaws
If you have multiple storage freezers, put your cells in the least-used one
Always do a positive control transformation to gauge efficiency
Buy new competent cells (yes, they cost money, but so does your time)
Try the homemade competent cells (rubidium chloride) method —many get better results than with the classic CaCl₂ prep
II. 3. Transformation Protocol
Small details can ruin your transformation:
Large volumes heat unevenly during heat shock; don’t go above 50 µL
Ensure your heat shock is actually at 42 °C
If using a heating block, put a bit of water in the well or use a PCR machine for precise 45-second pulses
After heat shock, recover cells in SOC or LB shaking for a full hour at 37 °C (if your lab says SOC isn’t important – it often is for difficult ligations; several folks swear by using SOC to boost colony count)
Plate as much of the transform mix as possible
If your vector is religating, use Alkaline phosphatase (CIP) to de-phosphorylate cut ends.
If you do CIP, gel-purify your vector afterward to remove every trace of the enzyme and any CIP-cleaved crap.
Pro tip: Your CIP can be too effective. I.e., it can dephosphorylate your insert ends as well. If that’s the case, switch to Antarctic Phosphatase (NEB), which is heat-inactivated and gentler on your DNA.
Some people simply use a 1:3 or 1:5 vector:insert ratio and a quick-ligation kit (5–15 min at room T°C) works without CIP, because the insert outcompetes vector recircularization.
II. 4. Insert Too Large or Toxic
If your insert is >5 kb or contains something bacteria don’t like (toxic genes, strong promoters), cloning can fail in mysterious ways.
One undergrad tried inserting a 3.3 kb cassette into a ~6.9 kb construct—a ~10 kb final plasmid. Not huge, but can be borderline.
Large inserts may ligate less efficiently and transform poorly.
What you can do
Use high-efficiency (electrocompetent) cells—they can give 10x higher efficiency for big plasmids
Incubate ligations overnight at 16 °C
Verify the insert’s sequence—does it have cryptic sites or needs a different strategy?
Modern methods that can bypass some classical ligation headaches: Ligation-independent cloning (LIC) or Gibson Assembly are good alternatives.
II. 5. Zero-Insert Colonies (Vector Only)
Sometimes you get colonies. But all you’ve cloned is frustration.
Empty vectors or mutated inserts.
Vectors that weren’t fully digested can religate and transform better than the insert-containing plasmid—always run a digest check or gel-purify the cut vector
Use two different (not one) restriction enzymes to get non-complementary ends that won’t self-ligate (plus makes orientation correct by default)
There can be an unseen mismatch. E.g., if both insert and vector were cut with NotI, they can join, but with random orientation
Also, NotI ends are symmetric; if you accidentally only cut the vector and insert incompletely, the vector could religate with a filler piece
A ligation gel can confirm something happened, but it’s not foolproof
For tough cases, try a double digest (two enzymes) if possible, or add an extra enzyme cut to linearize any circular multimers
Genes with leaky expression of a toxin or even just high AT-rich sequences can be toxic and can fail to clone
Try cloning into a low-copy plasmid first or use a strain like Stbl3 or XL1-Blue that tolerates unstable sequences
II. 6. Real-life quotes from the cloning trenches
“I’m pretty much at the ‘have you tried crying or an exorcism?’ stage of troubleshooting.”
After countless failed cloning attempts. Don’t lose hope – we’ve all been there, and yes, sometimes a good cry helps!
“Don’t worry about the SOC. I tried side by side next to LB, little difference. It’s all about the competent cells.”
Conflicting opinions on SOC aside, everyone agrees competent cells quality is crucial. If things suddenly stop working, suspect the cells first.
“Also, something I’ve been told: take sterile scissors and cut the pipette tip whenever pipetting cells – a wider bore reduces shearing.”
Great tip for delicate transformations or if using small tips. Avoids shredding your precious bacteria.
“Recommend TA cloning the PCR product and then digesting for cloning into desired plasmid.”
When direct cloning of a PCR product with restriction sites fails, an old-school solution is to:
First clone the PCR product into a TA vector (if you used Taq it has that extra A overhang) or a blunt vector
Sequence it
Subclone via restriction
It’s more steps but can rescue a project on a deadline.
II. 7. Summary of Cloning & Transformation Fails and Fixes
Part III. Miniprep Miseries: Where’s My DNA?!
Scenario: After you finally get a clone, you grow an overnight, do a miniprep, and… yield is terrible. Or the DNA is weirdly smeary.
Let’s address some miniprep nightmares:
III. 1. Low Yield from High-Copy Plasmid
By all logic, high-copy plasmids in healthy E. coli culture should give lots of DNA. If you get <20 ng/µL, something’s off.
One colleague on ResearchGate described:
“Using NEB Monarch kit, 1.5 mL culture, rarely get >20 ng/µL… tried Qiagen’s kit, still low.”.
They even suspected the ampicillin had gone bad, letting plasmid-free cells overtake the culture.
Here’s how to troubleshoot:
Culture conditions
In the last case, the problem was growing a 4 mL culture in a tight 15 mL tube at only 30 rpm
Grow in a flask or tube with proper shaking (200–250 rpm)
Loosen the cap or use a vented one for airflow
Even 3 ml, 12-hour cultures are fine for high-copy plasmids—only if well aerated
Otherwise, scale up to 10 mL in a 50-100 mL flask for better yield
Fresh antibiotic = happy plasmid
Plasmid-bearing cells produce β-lactamase that inactivates Amp. Ampicillin in particular degrades within a day in culture.
Use carbenicillin (more stable)
Ensure you harvest at ~16 hours, not more
Strain matters
Standard DH5α is slow-ish growing.
And “high copy” plasmids can have elements that reduce yield (e.g. certain origins, or large size ~10 kb+).
Use 5 mL culture instead of 1.5 mL to fix low density
III. 2. Column Overload and Other Kit Quirks
Paradoxically, using too many cells can clog a miniprep column, giving lower yields.
If you see a lot of viscous goop during lysis or the column flow-through is slow, you might have too big a pellet.
Culture 3 mL for high copy
Culture 10 mL for low copy
Resuspend cells in Buffer P1 + RNase—vortex until no clumps!
Ensure buffer P2 and N3 (or equivalents) are mixed well—incomplete lysis or neutralization leaves DNA in the pellet
4 pro tips and one myth
Pro tip 1: Ethanol in wash buffers can evaporate if the kit bottles sit around loosely capped. This can destroy yields
Pro tip 2: Heat your elution buffer to ~50 °C and let it sit on the column for a minute before spinning—warm elution improves DNA recovery from the silica
Pro tip 3: If yields are borderline, elute in two rounds: 2×25 µL instead of one 50 µL
One myth is that running the flow-through again increases yield—tests show it doesn’t do much
III. 3. Plasmid DNA but Make it Smear
Does your miniprep DNA shows a blurry smear (not a nice band) on the gel? It’s likely genomic DNA contamination.
The reason? Vortexing after adding lysis buffer (P2) shears gDNA into the lysate.
Instead:
Only vortex after P1
Only gently invert to mix P2
Don’t skip the neutralization step
Endonuclease-rich strains can chop DNA and also cause smears (wild-type E. coli or contaminated cultures)
Luckily standard (DH5α and TOP10) strains are endA1-negative. Just don’t use random strains.
III. 4. Case: Stubborn Low-Copy Plasmids
Low-copy plasmids are usually BACs, or big vectors with copy number ~5–20 per cell.
Seasoned researchers will tell you that the solution is simple.
You have two options:
Either scale up your culture volume
Or use specialized kits (Maxiprep or modified protocols with chloramphenicol amplification)
Addgene suggests doubling the culture volume for each halving of copy number.
Part IV. Which kits to buy? (company-specific insights)
This section gives candid feedback on brands to help you make better purchases.
IV. 1. Polymerases
Taq is good for routine screening—cheap and cheerful.
But for sequence accuracy or >2 kb, use Phusion from NEB.
High-fidelity enzymes (Phusion, Q5) often need optimization. They prefer GC Buffers for GC-rich targets, and you must use their recommended annealing T°C calculation.
For tough templates KAPA HiFi is another favorite (some find it more forgiving with long amplicons)
Pro tip: Polymerases vary in proofreading—so if you get nonspecific junk with one enzyme, switching brands can magically solve it
IV. 2. Ligation Kits
NEB’s Quick Ligation Kit allows 5–10 minute ligations—great for routine cloning
But for tough ones (big inserts, blunt ends), PIs often swear by the classic T4 DNA Ligase overnight at 16°C
Some have noted certain quick ligase buffers can give more background; if in doubt, do a side-by-side with a traditional ligation
IV. 3. Competent Cells
NEB’s DH5α equivalent (NEB 5-alpha) or high-efficiency cells come tested at >10^8 cfu/µg
Thermo’s XL10-Gold or Agilent’s XL1-Blue are specialized for difficult cloning (e.g. XL1-Blue has tetracycline resistance and tolerates repeats)
For routine plasmids, many like Top10 or Stbl3 for routine plasmids
Unstable plasmids? NEB Stable
Remember: Commercial cells can lose efficiency if thawed improperly. Thaw on ice. Five minutes. Gentle flick to mix DNA. Cold-shock after heat shock.
IV. 4. Miniprep Kits
Qiagen is the old stalwart: reliable, with RNase A included (don’t forget to add it!)
Some folks feel Qiagen columns bind ~20 µg max, so if you overshoot that, you don’t get more
NEB’s Monarch kit has nice features (colored buffers, claim of higher yields for some)
A user in our research found Monarch yielded same as Qiagen in their scenario, so performance is comparable
If yields are low with both, it’s likely your technique or plasmid, not the kit
Zymo Zyppy mini preps are super fast with no centrifuge needed for some steps, but they may shear off a percent of yield for speed
If one kit fails you, try another; some column silica or buffers work better for weird plasmids
Part V. Data & Presentation Pitfalls
You need this section—because getting the DNA is only half the battle.
V. 1. qPCR Shenanigans
Ever see amplification curves dip in late cycles?
Scenario
One user reported SYBR qPCR curves dropping around cycle 30.
It turned out to be an instrument quirk—the machine might’ve struggled with high fluorescence, giving a false drop.
The dissociation (melting) curves were normal, suggesting no real DNA loss.
Lesson
Don’t trust weird qPCR kinetics blindly.
Funky amplification curves + OK melt curve and gel = likely a detection artifact
I.e., the machine’s inability to process high signal.
In doubt?
Run a gel of the qPCR product to confirm presence/size
Contact tech support—sometimes it truly is a machine or software issue
V. 2. Inconsistent qPCR replicates
It’s tempting to throw out an outlier replicate.
The good news is: Yes, you can omit outliers if you have a rationale (e.g. machine error, obvious pipetting mistake).
Just be transparent.
To identify outliers quantitatively, you might calculate the coefficient of variation (CV) of replicates.
An Excel formula you can copy
To get CV% of your triplicate Cq values that are in cells B2, C2, D2, use:
=STDEV.S(B2:D2)/AVERAGE(B2:D2)
Many aim for CV <1% in qPCR. Anything much higher might mean an outlier.
GraphPad Prism also has built-in stats to flag outliers (e.g. Grubbs’ test). But use with caution—biological variation is real, and not every outlier is “invalid”. Sometimes inconsistent qPCR = you need to optimize primer efficiency or template prep.
V. 3. Huge error bars
Your qPCR or experiment has one condition with wild variability?
If you missed our article on data representation, you can read it here: “How to Choose Biology Figures Like an Artist”.
But the quick tip here is to use a different visualization
Change your bar graph with ±SD to plotting each replicate value (Prism’s “scatter dot with error bar”) → shows the spread without implying too much confidence in the mean
Log-transform skewed data (like copy numbers) to make error bars more interpretable (they become symmetric on a log scale)
V. 4. Excel gremlins
Lastly, a PSA:
Excel might autocorrect gene names (TURN1 becomes “Turn1” or dates)
Turn off auto-format and double-check that “Sep 2” in your data isn’t misread as a date
Use cell , use cell references carefully and document the formula logic in a worksheet note—many grad students were haunted by Excel sheets they don’t understand months later
Conclusion
Don’t rage-quit your PCR yet. Someone must’ve been in your shoes before.
Print this guide. Hang it on your bench. Share it with a colleague.
…and stay tuned for more content from Wildtype One.