A lab buyer is building a comparison sheet for one peptide from three suppliers. The published HPLC purities are 98.2%, 98.7% and 99.3%, and the spreadsheet puts them in that order with a green cell for the highest. Before the order goes out, a colleague in the analytical group asks one question: how many of those digits are real?
The honest answer is probably fewer than the spreadsheet suggests. A purity result has a spread around it, called its measurement uncertainty, and for an HPLC area-percent value on a peptide that spread is often wider than the gaps between the three numbers. This article explains where the spread comes from, roughly how large it tends to be, and how to use purity figures without over-reading them.
What measurement uncertainty means
Every measured value is an estimate. Measurement uncertainty describes the range within which the true value is reasonably expected to lie. The internationally used framework, the Guide to the Expression of Uncertainty in Measurement (usually called the GUM), sorts contributions into two types. Type A contributions are evaluated statistically from repeated measurements. Type B contributions are estimated in other ways, such as from calibration certificates, instrument specifications or experience. Independent contributions are combined as the square root of the sum of their squares to give a combined standard uncertainty. That is often multiplied by a coverage factor, commonly k = 2, to give an expanded uncertainty that corresponds to a coverage of roughly 95%.
Uncertainty is not the same as error. An error is the difference between one result and the true value, which is never known exactly. Uncertainty describes how well the method can pin the value down. A lab that reports it is showing that it understands its method, not admitting to a fault.
Where the spread in an HPLC purity figure comes from
An HPLC purity for a peptide is normally an area percentage: the main peak’s area divided by the total integrated area at one detection wavelength. The reverse-phase HPLC purity article describes the method. Several stages add variation to that ratio.
Repeatability of the sample load
Load the same solution several times and the peak areas will differ slightly. On a well-maintained system the relative standard deviation of the main peak is typically small, but minor peaks near the limit of quantitation vary much more in relative terms, and their total is what sets the purity figure.
Integration decisions
Where the baseline is drawn, how a shoulder is split from the main peak, and which small peaks count as noise are all choices. Two competent analysts, or one software package with two parameter sets, can report different areas from the same data file. The effect grows as peaks tail or overlap.
Detector response
Area percent assumes every component gives the same signal per unit mass. At low UV wavelengths around 214 to 220 nm, where the peptide bond dominates absorbance, this is a reasonable approximation for related peptide impurities. At 280 nm it depends on aromatic residues, and an impurity missing a tryptophan or tyrosine can be badly under-counted.
Incomplete separation
Anything that co-elutes with the main peak is counted as product. This contribution is one-sided: it can only push purity up, never down. It does not appear in repeatability data at all, which is why a result can be very precise and still biased.
How big the band usually is
The size depends on the method and the sample, and a laboratory can only state it properly after evaluating its own method. As a general guide, a well-resolved, symmetric main peak measured at a low UV wavelength might give a purity with an expanded uncertainty of a few tenths of a percentage point. Tailing peaks, shoulders or crowded impurity profiles can widen that toward a full percentage point or more, and co-elution adds an unknown upward bias.
The following illustration shows how contributions combine. The numbers are invented for the example, not taken from any real method.
| Contribution (illustrative) | Standard uncertainty, % points |
|---|---|
| Repeatability of the load | 0.10 |
| Integration choices | 0.20 |
| Response-factor differences | 0.15 |
| Combined (root sum of squares) | about 0.27 |
| Expanded, k = 2 | about 0.54 |
With a band of roughly ±0.5 points around each figure, the three supplier results in the opening example overlap. The 98.7% and 99.3% values cannot be told apart, and even the lowest and highest are close enough that the order could change on another day.
Other reasons two figures differ
Uncertainty within one method is only part of the picture. Figures from different laboratories often come from different columns, gradients, wavelengths and integration rules, so they were never on the same scale. Our article on why suppliers report different peptide purity covers those method differences. Material variation adds another layer. At Battle Born, Battle Born tests every product by independent reverse-phase HPLC and publishes that result product by product. Testing is per product, not per production batch, so a buyer should also allow for normal batch-to-batch variation when relating a published figure to a particular vial.
A practical checklist for comparing purity figures
- Regard differences of a few tenths of a percentage point as ties unless both labs used the same method and state their uncertainty.
- Check that the detection wavelength is stated, and prefer results at a low UV wavelength for peptide purity.
- Look at the chromatogram itself: peak shape, baseline and resolution of neighboring peaks matter more than the final digit.
- Ask whether identity was confirmed by a separate technique, since purity alone does not establish what the main peak is.
- Note what the purity figure does not cover, such as water, counter-ions and non-peptide content.
- Follow up only on gaps of several percentage points, or on results that fall below the specification you need.
Price comparisons have the same trap. A decimal-point difference in purity is rarely worth a real difference in cost, as discussed in how to compare research peptide prices.
Questions
Why do most peptide purity reports not state an uncertainty?
Evaluating uncertainty properly takes method validation work that many research-grade testing arrangements do not include. A missing value means the precision is unstated. It does not mean the result is wrong.
Is a more precise number more accurate?
No. Precision describes how closely repeated results agree. A method can be very precise and still biased, for example when an impurity co-elutes with the main peak.
What does k = 2 mean?
It is the coverage factor applied to the combined standard uncertainty. For approximately normal distributions, k = 2 gives an interval expected to contain the true value with about 95% confidence.
When is a purity difference worth raising with a supplier?
When it is larger than the plausible combined uncertainty of both results, typically several percentage points, or when a result falls below the specification your work needs.
Research use only. All products supplied by Battle Born Peptides are laboratory reference materials for in-vitro research and analytical use by qualified professionals. They are not drugs, foods, dietary supplements, cosmetics or medical devices; they are not approved by the FDA or any other regulator for use in humans or animals; and they are not intended to diagnose, treat, cure, mitigate or prevent any disease, or to affect the structure or any function of the body of humans or animals. Nothing in this article is preparation, handling or dosing guidance. See our full research-use terms.