Common Mistakes in Fragment Morphology: Avoiding Pitfalls for Improved Structure-Activity Relationships

Reading a gel image or an electropherogram looks straightforward, and that is exactly why it goes wrong so often. The most damaging errors in DNA fragment analysis are not clumsy pipetting; they are confident misinterpretations of ambiguous data. A band that is not a fragment, a size read beyond the ruler, an artefact called as a real allele: these mistakes produce clean-looking, entirely wrong conclusions. This article catalogues the most common interpretation pitfalls in nucleic-acid fragment analysis and how to avoid each one.

Want expert help putting this into practice? FragmentMorphology can guide you through it.

Mistaking Artefacts for Real Fragments

The single most frequent error is treating every band or peak as a genuine fragment. Separation platforms generate predictable artefacts, and calling them as results corrupts the interpretation.

• Primer dimers appear as small products below the true amplicon; they are real DNA but not your target.
• Stutter peaks in repeat analysis sit one repeat unit below the true allele and must not be counted as a separate allele.
• Pull-up (bleed-through) in multi-colour capillary work produces a false peak in one colour channel directly beneath a tall peak in another; it is an optical artefact, not a fragment.
• Split peaks show one fragment as two closely spaced peaks, usually from incomplete terminal base addition.

The remedy is to learn the artefact profile of your specific assay and check every questionable peak against it before assigning meaning.

Sizing Outside the Standard's Range

Related: Fragmentmorphology Tips and Strategies for Better Writing.

A pervasive error is reading a size for a fragment that migrated beyond the reference bands. Sizing is interpolation between known points, and interpolation is only valid within the range spanned by the standard.

When a fragment runs faster than the smallest ladder band or slower than the largest, its size is an extrapolation, and the sizing curve is unreliable there. On gels this often happens when fragments run near the well or off the far end; on capillary systems it happens when a peak falls outside the internal standard's fragments. The fix is to choose a standard that fully brackets the expected fragments and to flag, rather than report, any peak outside that range.

Reading Migration Distance Linearly

Beginners frequently assume that a fragment halfway down the gel is half the size of the largest. It is not. The relationship between fragment size and migration distance is approximately logarithmic, so equal distances do not represent equal size differences.

Estimating sizes by linear eyeballing badly overestimates small fragments and underestimates large ones. The correct approach is to build a calibration curve, plotting the logarithm of known ladder sizes against migration distance, and read unknowns from that curve, or let software fit it. Trusting a mental linear model is a reliable route to sizes that are wrong by large margins.

A related error is comparing migration between two separate gels as if a band at the same height means the same size. Small differences in gel percentage, buffer age, voltage, and run time shift migration between runs, so a fragment can sit at visibly different positions on two gels and still be identical. If two samples must be compared for size, run them on the same gel in adjacent lanes with a shared ladder; cross-gel comparison by position alone invites false conclusions that a fragment has changed size when only the run conditions differed.

Ignoring Band and Peak Shape

See also: Fragmentmorphology - Expert Advice for Better Writing.

Analysts often focus on where a band sits and ignore what it looks like, but shape carries information about data quality.

• Smeared bands usually indicate degraded DNA or overloading, not a real spread of fragment sizes.
• Wavy or shifted bands point to excess salt or uneven field, meaning the migration, and therefore the size, is unreliable.
• Broad, saturated peaks on a capillary trace signal over-injection; the true size and any minor neighbouring peaks may be obscured.

Interpreting the position of a malformed band as a precise size ignores the warning that the run itself is compromised. Assess shape before assigning any number.

Confusing Intensity With Identity

Band or peak height is a rough measure of DNA quantity, not of fragment identity, yet the two are often conflated. A faint band is not necessarily a smaller or less important fragment, and a bright band is not necessarily the correct product.

Two common consequences follow. First, analysts sometimes dismiss a faint but genuine band as noise, missing a real fragment. Second, they treat the brightest band as the target when it is actually an abundant artefact such as primer dimer. Quantity and identity are independent: establish what a fragment is from its size and expected pattern, and use intensity only as a secondary, quantitative observation, ideally normalised against controls.

Skipping the Controls and the Expected Pattern

The final common mistake is interpreting a lane in isolation, without reference to controls or to the pattern the experiment predicts. A restriction digest, for example, should produce a specific number of bands whose sizes sum to the known total; if they do not, the digest was incomplete or the map is wrong, and no single band should be over-interpreted.

Always run and read the controls first: a no-template control reveals contamination, and a positive control confirms the run worked. Then compare the sample against the expected pattern rather than reading it fresh. Interpretation is a comparison against expectation, not a description of whatever bands happen to appear.

A subtle version of this mistake is confirmation bias in the other direction: seeing the pattern you expected and stopping there. If a digest happens to produce roughly the right number of bands at roughly the right places, it is tempting to declare success without checking that the sizes actually sum correctly or that a faint extra band is not lurking. The discipline that guards against both over-reading and under-reading is the same: state what the experiment predicts before you look, then check the data against that prediction point by point, treating any deviation as information rather than an inconvenience to explain away.

Avoiding these pitfalls comes down to a single disciplined habit: never assign meaning to a band or peak until you have confirmed it is a real fragment, that it lies within the standard's range, that its shape is sound, and that it fits the pattern the experiment predicts. The reviewers at FragmentMorphology find that most interpretation errors trace back to skipping one of those checks in the rush to a conclusion, and that slowing down at the reading stage, where a mistake is invisible until it has propagated downstream, is the cheapest quality improvement available in the entire fragment-analysis workflow.