Current turnaround: 21 days from receipt of your samples. Due to high demand, HPLC-UV peptide testing and Certificates of Analysis are taking longer than usual.

Independent Research Testing Center · 875 Innovation Drive, Cambridge, MA 02139, USA For all independent testing inquiries, email admin@novagenlaboratories.com
How a lab works

Method Validation: Why the Method Is Tested Before the Sample

A result is only as defensible as the method behind it. Validation is the work done beforehand to show what a method can and cannot measure.

Novagen Analytical Labs

A method is tested before the sample because a measurement means nothing until you know how the measuring process behaves. If a laboratory has not shown that its method is accurate, repeatable and able to tell the target compound from everything else, then a purity figure produced by that method is just a number with no evidence behind it. Analytical method validation is the evidence. It is done first, on known materials, so that when a real sample is run later, the result can be defended.

This is easy to overlook because it is invisible on a certificate. The certificate shows the sample result. The validation work sits behind it and decides how far that result can be trusted. What follows explains each part of that work in plain terms, without the jargon.

Why the method comes before the sample

Think of weighing something on a scale. Before you trust the reading, you would want to know the scale reads zero when empty, gives the same weight if you place the object on it twice, and gives the right weight for an object of known mass. You test the scale before you test the object.

An analytical method is the same idea with more moving parts. A peptide purity method involves a column, a solvent gradient, a detector, a reference standard and a calculation. Each of those can introduce error. Validation is the structured way of finding out how large that error is, and under what conditions it stays small.

The public guideline most laboratories use as a framework is ICH Q2, titled Validation of Analytical Procedures. It sets out the characteristics that should be considered and describes them generically. It does not hand every method the same checklist. The characteristics tested depend on what the method is for. A method that identifies a compound needs different proof from one that quantifies an impurity.

Accuracy: is the number close to the true value?

Accuracy is the closeness of a measured result to a value accepted as true. The difficulty is obvious: with an unknown sample, nobody knows the true value. So accuracy is tested on material where the answer is known.

The usual approach is a recovery study. A known amount of the target compound is prepared, often by spiking a known quantity into the sample matrix, and the method is asked to measure it. If the method reports back close to what was put in, recovery is good. If it consistently reports too little or too much, the method has a bias that has to be understood or corrected.

This is why reference standards matter. Accuracy can only be judged against something whose identity and content are independently established. Without that anchor, a method can agree with itself perfectly and still be wrong.

Precision: does it give the same answer twice?

Precision is the agreement between repeated measurements of the same material. It says nothing about whether the answer is right. A method can be very precise and consistently off.

Validation separates precision into levels, because different sources of variation appear at different scales.

  • Repeatability. The same analyst, the same instrument and the same conditions over a short interval. This is the smallest variation the method will show.
  • Intermediate precision. Variation within one laboratory across different days, different analysts and different equipment. This is closer to real life, because a result reported in March should agree with one reported in June.
  • Reproducibility. Agreement between different laboratories. This applies when a method is transferred or compared across sites.

Intermediate precision is the one that quietly matters most. A method that only behaves on one good morning with one careful analyst is not a method. It is an anecdote. Testing across days and operators is what shows the method does not depend on who is running it.

Specificity: can it tell the target from everything else?

Specificity is the ability to measure the target compound unambiguously in the presence of other things that are expected to be there. For a peptide, those other things include synthesis by-products, degradation products and components of the sample matrix.

This is the characteristic most closely tied to purity. A purity figure is a statement about what share of the material is the target. If a method cannot separate the target from a closely related impurity, the two blur into one peak and the reported purity is too high. The method has not seen the impurity. It has absorbed it.

Specificity is commonly demonstrated by showing that the target peak is resolved from known related substances, and by stressing a sample deliberately, for example with heat, acid or oxidation, to generate degradation products and confirm they do not hide under the main peak. This connects directly to peak purity and coelution, where the same concern is seen on a real chromatogram.

Where one method cannot achieve full specificity on its own, a second, different method is sometimes used alongside it. That is the reasoning behind pairing chromatography with mass spectrometry.

Linearity, range and the lower limits

Linearity is the ability of a method to give a response that is proportional to the amount of compound present, within a given interval. Double the concentration and the signal should roughly double. If it does not, a calculation that assumes proportionality will produce the wrong number.

This is tested by preparing a series of solutions at different known concentrations and checking how the response follows them. The range is the interval between the lowest and highest concentration over which the method has been shown to perform with acceptable accuracy, precision and linearity. A result outside that range is not covered by the validation, and it should not be reported as if it were.

Two related characteristics describe the bottom of the range. The detection limit is the lowest amount that can be detected, though not necessarily measured as an exact value. The quantitation limit is the lowest amount that can be measured with acceptable accuracy and precision. Our note on limit of detection and quantitation covers the difference in detail. For purity work it matters because a small impurity near or below the quantitation limit cannot be given a trustworthy number.

Robustness: does it survive small changes?

Robustness is a measure of how little a method is affected by small, deliberate changes in its operating conditions. In liquid chromatography, those might be a slight shift in mobile phase composition or pH, a small change in column temperature, a different flow rate, or a different batch of column packing.

The reason to test this is practical. Real laboratories never reproduce conditions perfectly. A solvent is mixed slightly differently, a column ages, the room is warmer. If a small change in any of those flips the result or merges two peaks, the method is fragile, and that fragility needs to be known and controlled.

Robustness is typically studied during method development, before the formal validation, so that the method that gets validated is one already known to tolerate ordinary variation. A method that is not robust is one reason the same sample can return different numbers in different places. Our piece on why labs report different purity explains how method differences, and not only sample differences, produce that spread.

System suitability: checking the method still works today

Validation is done once. It does not guarantee that the instrument is behaving on a given afternoon. System suitability is the routine check that fills that gap.

Before and during a run, the analyst injects a reference preparation and checks that the system meets predefined performance criteria. Typical things examined include whether the target peak is adequately separated from its nearest neighbour, whether the peak shape is acceptable, and whether repeated injections of the same standard agree closely. If the checks fail, the run is not used. The samples are not reported on a system that has not shown it is performing as validated.

The two ideas are complementary. Validation shows the method is capable. System suitability shows it is capable right now.

What validation does not do

A validated method is not a guarantee, and the limits are worth stating plainly.

  • It covers a defined scope. A method validated for one compound, one matrix and one concentration range is not automatically valid for another.
  • It does not describe the batch. It shows the method measures reliably. It says nothing about whether the sample submitted represents a wider lot. That question is covered in single vial versus batch testing.
  • It does not turn purity into safety. Purity is composition. A well-validated purity method still reports composition and nothing more.
  • It is not permanent. A significant change to the method, the instrument or the type of sample can require some or all of the validation to be repeated.

What validation does give is the thing a single number needs in order to be defensible: a known level of confidence about how far the number can be trusted, and the conditions under which that holds.

What this means when you read a result

A Certificate of Analysis reports what was measured in the sample submitted. The method named on it is the method that was validated, and the validation is what stands behind the figure. If a certificate gives a purity number and names no method at all, there is nothing to check the number against. Our walkthrough of how to read a peptide Certificate of Analysis covers which fields to look for.

Novagen is a third-party analytical laboratory. It reports what was measured in the sample submitted, and it publishes a verifiable certificate register at novagenlaboratories.com. Current quoted turnaround is 21 days from receipt of samples. Questions can be sent to admin@novagenlaboratories.com.

Common questions

What is analytical method validation?

It is the documented process of showing that an analytical method is suitable for its intended purpose. A laboratory tests the method itself, using known materials, to establish how accurate, precise, specific and robust it is, and over what concentration range it works. This is done before the method is used to report a result on a real sample.

What are the main parameters of method validation?

The characteristics described in the ICH Q2 guideline are accuracy, precision (including repeatability and intermediate precision), specificity, linearity, range, detection limit and quantitation limit. Robustness is closely related and is usually studied during method development. Which ones apply depends on what the method is meant to measure.

What is the difference between accuracy and precision?

Accuracy is how close a result is to the true or accepted reference value. Precision is how closely repeated measurements of the same sample agree with each other. A method can be precise and consistently wrong, or accurate on average and scattered. Validation tests both, separately.

What is the difference between method validation and system suitability?

Validation is done once, or again after a significant change, to show that the method is fit for purpose. System suitability is a set of checks run with the instrument on the day, alongside the samples, to show that the method is still performing as it did when validated. One proves the method. The other proves the method is working today.

Does a validated method guarantee the result is right?

No. Validation shows that a method performs reliably within defined conditions and limits. It does not guarantee that a particular sample was prepared well, stored well or representative of a batch. A validated method makes a result defensible. It does not make the result a statement about anything beyond the sample that was tested.