HPLC and SDS-PAGE Purity: Which Number Should You Trust?

Updated 2026-09-28

HPLC and SDS-PAGE measure different properties. Understanding what each purity figure captures helps developers compare recombinant proteins fairly.

Two Methods, Two Questions

Purity is not a single property, so a single number cannot describe it. HPLC methods such as size-exclusion or reversed-phase chromatography separate species by size or hydrophobicity in solution and report the proportion of the main peak. SDS-PAGE separates denatured proteins by molecular weight and reports the proportion of the main band. The two methods disagree because they interrogate different behaviours: one observes native aggregation and hydrophobicity, the other chain length and fragmentation. A high value from one method does not compensate for an unmeasured property in the other.

What Each Figure Captures

An HPLC purity value is sensitive to aggregates, fragments and isoforms that differ in surface properties, and it measures material that may not enter a gel. An SDS-PAGE value reflects the staining intensity of separated bands and is comparatively insensitive to non-covalent aggregates that dissociate under reducing conditions. Neither method detects the other's blind spots. A low-molecular-weight impurity may dominate a gel while being poorly resolved by a size-exclusion method, and the reverse can also hold.

Reading Them Together

Report both figures with their methods and conditions, and treat the pair as a description of the material rather than a single verdict. Where a specification matters, name the method explicitly, for example size-exclusion HPLC or reducing SDS-PAGE, so comparisons across suppliers rest on the same basis. If only one purity figure is quoted, ask which method produced it and whether the other was measured. The question is not which figure to prefer, but what each was designed to detect.

The third number almost nobody asks for

Purity is only one of three properties that determine whether a recombinant protein will behave in an assay, and it is the one suppliers compete on because it is the easiest to quote. The second is identity, which a purity figure says nothing about at all: a preparation can be 98% pure and still be 98% of the wrong protein, which is why identity confirmation by mass spectrometry, peptide mapping or a specific immunoblot is a separate line on a well-constructed CoA. The third is biological activity or, at minimum, an activity-correlated measure such as binding to a specific antibody. For a calibrator antigen the third property is the one that determines whether your standard curve means anything, and it is frequently absent from the documentation. Asking for it is not unreasonable, and the answer tells you whether the supplier characterises the protein or only purifies it.

Choosing a method-specific specification

The reason to name a method in your specification is that 'purity 95%' is not a specification at all until you say by what. If your concern is aggregation in solution, specify the chromatographic method and the limit for the main peak. If your concern is fragmentation or a truncated construct, specify a reducing gel and the limit for the main band. If your concern is both, as it usually is, specify both and report both, accepting that they will rarely agree numerically — and say so internally, because a reviewer who sees two different purity figures for one lot will otherwise assume one of them is wrong. Writing the specification in method-specific terms also makes supplier comparison possible for the first time, since two figures obtained by the same named method are at least talking about the same property.

Ask a technical question