Develop methods you can carry into QC.

Method development usually leaves its evidence in several places: preparation notes, sequence files, processing settings and a final procedure written afterwards. In Seal the study record is made as the work happens. The protocol defines the preparation steps, required inputs, sample links, calculations and review; each run keeps its own values, observations, evidence and history. The development evidence stays with the method into validation and routine use, and a later review starts from what was done rather than from a reconstruction of it.

Describe the method and Neil proposes the protocol: its steps, fields, calculations and review. Your team tests the workflow with representative runs and defines the review and approval requirements for its intended use, then reuses the protocol for every study that follows. The structure fits the technique: chromatographic, electrophoretic, spectroscopic and biological assays each record their own preparations, responses and review requirements.

A later protocol revision does not rewrite earlier runs. Each study keeps the version it followed, so a comparison across studies shows where the method itself changed. The electronic lab notebook describes protocols, calculations and experiment records in more detail.

The fictional study AD-041 varies buffer pH (6.8, 7.0 and 7.2) and column temperature (25, 30 and 35°C) around a nominal condition of pH 7.0 and 30°C, and measures critical-pair resolution against a study criterion of at least 1.5. There is one observation per condition, recorded as AD-041-01 to AD-041-09, with the same sample preparation, column lot and processing method. No replicate variability or full validation dataset is supplied.

ICH Q14 describes robustness as a procedure’s capacity to meet its expected performance criteria during normal use, tested by deliberate variation of its parameters. For most procedures it is evaluated during development and need not be repeated in validation, and its outcome should be documented and reflected in the analytical procedure control strategy.²

Eight passing results do not make a range. Each is a single observation, and the conditions between the tested points are untested. ICH Q14 describes a method operable design region as combined parameter ranges within which the procedure is shown to be fit for its intended purpose.² These nine observations do not show that; they show where to look next.

That makes the failing condition the most useful result in the study. A summary report tends to leave it out. In Seal, AD-041-09 remains a record with its own preparation, sequence and processing context, and the next study starts from it.

Neil’s draft for AM-041 recommends investigating the edge before approving the range. Eight conditions meet the selected resolution criterion and one does not, which suggests sensitivity to the tested pH and temperature settings. Repeatability, other performance characteristics and the conditions between the tested points remain open.

The proposed follow-up has four steps. Review the original sequence, integrations, sample preparation and instrument conditions for AD-041-09, retaining the original result and any reprocessing history. Propose independent preparations at the nominal and failing conditions, with run order and replication justified in the study plan. Assess additional conditions near the observed loss of separation, with the relevant column lots and sample matrices in the risk assessment. Then reassess the proposed parameter controls against the intended use and the remaining performance requirements, and submit any revised range for scientific review.

Keeping the original result matters. FDA expects each result from reprocessed chromatography to be retained for review, not only the final one.³ Neil keeps its findings tied to the supplied records and distinguishes observations from hypotheses. It does not decide that a result is invalid, that a range is acceptable or that a method is ready.

Neil works within the permissions of the person asking, and customer data is not used to train AI models. In line with the EU’s draft GMP Annex 22 on AI, Neil is not used in GMP execution. It helps set up configuration, which your team verifies and releases under change control. See how Neil fits Annex 22 and more about Neil.

Start with what the method must measure. An analytical target profile describes the procedure’s intended purpose, the product attributes it measures and the performance characteristics it needs, with their criteria; ICH Q14 treats it as the basis for development and for the validation criteria, and keeps it through the lifecycle.⁴ In Seal the target profile is a record linked to the product, matrix, reportable result and performance needed, and each development study states the question it answers.

Validation then tests the method against those criteria. ICH Q2(R2) asks for a validation protocol before the study, stating the intended purpose, the performance characteristics to be validated and their criteria, with a justification wherever development data or other prior knowledge is used. For an assay or a quantitative impurity test, the typical characteristics are specificity, range, accuracy and precision.⁵ A stability-indicating method also needs samples that contain the relevant degradation products, such as stressed samples from forced degradation studies.⁶

The readiness review gathers that evidence before the protocol is written: the target profile, the method definition and its control strategy, robustness and selectivity results, reference standards, calculations, sample stability and data integrity, with each unresolved gap visible. ICH Q14 expects the analytical procedure control strategy, including its system suitability test, to be defined before validation and confirmed after it.⁷

Your team defines the applicable performance characteristics, criteria and validation work. A robustness plot or a generated report does not establish validation by itself, and Neil’s gap review is a list of what is missing, not an approval. Validation studies are then linked to the method version they cover, so the evidence traces to what was actually validated.

The instructions are only part of the method. Carry the preparation details, standards, processing settings and known sensitivities into transfer: the samples, standards, lots, quantities, dilutions and preparation times; the instrument, column, sequence position and acquisition method; the original data, processing version, calculations and any reprocessing rationale; and the suitability outcomes, deviations, agreed criteria and scientific assessment. ICH Q14 sets the standard for the written procedure: enough detail for a skilled analyst to perform the analysis and interpret the results.⁷

Moving the method to QC does not complete a transfer. ICH Q2(R2) expects a transfer to another laboratory to include partial or full revalidation or a comparative analysis of representative samples, or a justification for doing neither; co-validation across sites can also meet it.⁸ EU GMP sets out the steps: a documented gap analysis of the original validation, a transfer protocol covering the methods, training, standards and samples and the acceptance criteria, investigation of any deviation, and a report of the comparative outcome.⁹ In the United States, the suitability of every testing method must be verified under actual conditions of use.¹⁰

In Seal the transfer is its own work, linked to the method version: the strategy and criteria, the receiving lab’s equipment, people and materials, the supporting studies and the receiving team’s review. Compare receiving-lab results before concluding the method is suitable there; a development package does not approve the method or complete a transfer. A compendial method is verified rather than transferred: link the USP, Ph. Eur. or JP method to the site’s verification and assess the impact when the compendium changes. The method lifecycle and technology transfer pages follow that work further, and the QC laboratory page follows the method into routine testing.