Documenting a method someone else can run

A method someone else can run is a record that survives your absence. The chromatogram is the result. The record is the column, the liquids, the programme, and the detector settings that made it. This page is the application of that idea to a life-science separation. What the result is allowed to mean sits in chromatography methods in life-science labs.

The lines that have to exist

Chromatography is defined by the phases and the movement. Write the stationary phase as a product identity: chemistry, particle size, pore size if it matters, internal diameter, length, and the lot if you have it. Two C18 columns with different lots can differ. A replacement later is a guess without the lot, and a fair attempt with it.

Write mobile phases as recipes a stranger can weigh or pipette. "Buffer A" is not a recipe. Solvent names, concentrations, pH, the acid, and the date and lot of a critical bottle are the recipe. If a pipette builds it, the volume is part of the method, and ISO 8655-1 is the background expectation for that class of device. Write the gradient as a table of composition against time, and also against column volume if you know the bed. Write flow, column temperature, injection volume, and the sample solvent. Write detection: wavelength, or excitation and emission, or mass range and polarity. Write dwell, measured or taken from the configuration you actually plumbed, because the table at the mixer is not the table at the bed.

Equilibration length, the standard, and the retention window close the loop. Without them, a colleague can run your gradient and have no rule for stopping when the system has drifted.

Affinity, ion exchange, and the HPLC assay are one campaign

A protein campaign often has an affinity step, an ion-exchange step, and an HPLC check. Document them as separate methods that name the fraction handed across. The affinity record needs the ligand, the metal, the bed volume, the linear velocity or the gravity head, the wash and elution, including imidazole if that is the competitor, and which fractions were pooled after which assay. The ion-exchange record needs anion or cation, pH, the salt programme, conductivity at elution, and the pool. The HPLC record needs the lines in the previous section. A single paragraph that merges all three will drop dwell or drop the imidazole concentration, and the next person will invent one.

Say what the pool was dissolved in before the next injection. Sample solvent is the usual omission, and it is the usual reason a transferred method suddenly fronts. Say what was not pooled. The flow-through that held activity is a result, not a detail.

If a sequence mass matters, point at the construct and, where it helps, at UniProt. If a mass-spectrometry identification was claimed, write the reporting level you actually met. HUPO is the community context for how strong that claim should be. A retention time is not a substitute sentence for it.

A transfer test

Hand the record to someone who did not watch you, on the same instrument first. They should not need to ask what "the usual flow" is. Compare their standard retention and area with the window. If they miss, the missing line is the defect in the document, not in the colleague. Then transfer to a second instrument if the laboratory has one. Expect to edit expected times for dwell. If you edit silently and do not write the second dwell, you have two methods and one title.

Store a raw file, not only a processed picture. Integration parameters change area. The picture with the baseline drawn is an interpretation. The raw trace plus the processing record is the method's output. Note the software version if processing is part of the number you publish.

Application in a shared instrument room

The useful test of this document is a Monday run by a person who has the file and the column, and no memory of Friday. Put the record where the instrument is, including the equilibration and the shutdown into storage solvent the insert names. A method that ends in high salt and does not say so leaves the next user a precipitated frit. Note pressure and temperature on the run sheet so a hot afternoon is visible data. If power fails often, write the restart: re-equilibrate, inject the standard, and only then continue. An unwritten restart is how half a batch is reported under a method that was not running.

Version the record when you change the acid, the dwell, or the window. Keep the old version with the data it produced. Editing the only copy makes last month's peaks impossible to defend.

Limits

A complete record supports a research repeat. It is not a validated diagnostic method and not a regulatory submission. Biosafety of the sample is still your institution's decision. Chemical hazards of the mobile phase belong in the record as handling notes, not as an afterthought in a chat message. Do not put secrets you cannot share in the only copy if a colleague is expected to run the separation safely.

Enquiry

When you ask for a column or a system to host a method, send the record: identity, lot if you have it, phases, gradient, flow, temperature, injection, detection, dwell, and the window. Ask for a match to chemistry, particle size, dimensions, pressure rating, and pH window. Use the scientific instruments catalogue and the quote request. If the method is a protein purification, the custom protein expression and purification reference is the enquiry context for discussing that method. ---