validate-analytical-method
关于
This Claude Skill automates validation of chromatographic analytical methods according to ICH Q2(R2) regulatory guidelines. It helps developers define the validation scope, establish specificity via forced degradation studies, and determine linearity, accuracy, precision, and detection limits. Use this skill to systematically generate the required validation protocols and data for regulatory submissions.
快速安装
Claude Code
推荐npx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/validate-analytical-method在 Claude Code 中复制并粘贴此命令以安装该技能
技能文档
Validate an Analytical Method
Formal validation of chromatographic analytical method following ICH Q2(R2) guidelines. Cover validation scope definition by method category, specificity/selectivity establishment, linearity and range determination, accuracy and precision assessment, LOD/LOQ and robustness studies for regulatory compliance.
When Use
- New chromatographic method developed and must be validated before routine use
- Compendial method being verified for suitability in specific laboratory
- Existing validated method has undergone significant changes requiring partial or full re-validation
- Preparing validation package for regulatory submission (NDA, ANDA, MAA, IND)
- Transferring method to new laboratory or instrument platform
Inputs
Required
- Developed method: Fully optimized, documented chromatographic method (column, mobile phase, gradient, detector, etc.)
- Method category: Assay of active ingredient, quantitative impurity test, limit test for impurities, or identification test
- Analyte reference standards: Primary reference standards with certificates of analysis, assigned purity
- Sample matrix: Representative samples including placebo/blank matrix for specificity studies
Optional
- Regulatory guidance: Specific regulatory requirements beyond ICH Q2 (e.g., USP <1225>, FDA guidance, EMA guidelines)
- Forced degradation samples: Pre-stressed samples (acid, base, oxidation, heat, light) if not yet prepared
- Validation protocol: Pre-approved protocol specifying acceptance criteria (required in GMP environments)
- Transfer package: Validating as part of method transfer? Originating lab's validation report
Steps
Step 1: Define Validation Scope per ICH Q2(R2)
Identify method category. Determine which validation parameters required.
| Parameter | Cat I: Assay | Cat II: Impurity (Quant) | Cat III: Impurity (Limit) | Cat IV: Identification |
|---|---|---|---|---|
| Specificity | Yes | Yes | Yes | Yes |
| Linearity | Yes | Yes | No | No |
| Range | Yes | Yes | No | No |
| Accuracy | Yes | Yes | No | No |
| Precision (repeatability) | Yes | Yes | No | No |
| Precision (intermediate) | Yes | Yes | No | No |
| LOD | No | May be needed | Yes | No |
| LOQ | No | Yes | No | No |
| Robustness | Yes | Yes | Yes | No |
- Classify method into one of four ICH categories based on intended purpose
- From table, identify all required validation parameters
- Define acceptance criteria for each parameter before beginning experimental work. Typical criteria:
- Linearity: R^2 >= 0.999 (assay), >= 0.99 (impurity)
- Accuracy: recovery 98.0-102.0% (assay), 80-120% at LOQ level
- Repeatability: RSD <= 2.0% (assay), <= 10% at LOQ level
- Intermediate precision: RSD <= 3.0% (assay)
- Draft validation protocol documenting all parameters, experimental designs, acceptance criteria
- Obtain protocol approval (in GMP environments) before beginning experimental work
Got: Approved validation protocol specifying method category, required parameters, experimental designs, pre-defined acceptance criteria.
If fail: Method category ambiguous (e.g., combined assay and impurity method)? Validate for most stringent category that applies. Consult regulatory guidance specific to submission type.
Step 2: Establish Specificity and Selectivity
- Prepare following solutions:
- Blank (solvent/diluent only)
- Placebo (matrix without analyte, e.g., excipients for drug product)
- Reference standard at working concentration
- Spiked placebo (matrix + reference standard)
- Forced degradation samples (if not already available)
- Perform forced degradation to generate potential degradation products:
| Stress Condition | Typical Treatment | Target Degradation |
|---|---|---|
| Acid hydrolysis | 0.1-1 N HCl, 60-80 C, 1-24 h | 5-20% |
| Base hydrolysis | 0.1-1 N NaOH, 60-80 C, 1-24 h | 5-20% |
| Oxidation | 0.3-3% H2O2, RT-60 C, 1-24 h | 5-20% |
| Thermal | 60-80 C, solid state, 1-7 days | 5-20% |
| Photolytic | ICH Q1B (1.2M lux-hours, 200 Wh/m^2 UV) | 5-20% |
- Inject all solutions, evaluate:
- No interfering peaks from blank or placebo at analyte retention time
- Degradation products resolved from main analyte peak (Rs >= 1.5)
- Peak purity confirmed by DAD spectral purity index or MS
- Calculate mass balance: assay + impurities + degradation products should account for 95-105% of initial content
- Document specificity results with chromatograms from all conditions
Got: Method demonstrated specific — no interferences from blank/placebo, degradation products resolved from analyte, peak purity confirmed, mass balance within 95-105%.
If fail: Degradation products co-elute with analyte? Method not stability-indicating. Return to method development to improve selectivity (adjust pH, gradient, column chemistry) before proceeding with validation.
Step 3: Determine Linearity and Range
- Prepare at least 5 concentration levels spanning intended range:
- Assay methods: typically 80-120% of target concentration
- Impurity methods: from LOQ to 120-200% of specification limit
- Dissolution: from 10-120% of label claim (or as needed for dissolution profile)
- Prepare each concentration level independent (not by serial dilution) for best practice
- Inject each level in triplicate (minimum duplicate)
- Perform linear regression of response (area or height) vs. concentration:
- Report slope, intercept, correlation coefficient (R^2)
- R^2 >= 0.999 for assay; R^2 >= 0.99 for impurity quantitation
- Evaluate residual plots:
- Residuals should be randomly distributed around zero with no systematic pattern
- Curved residual pattern indicates non-linearity — consider quadratic fit or narrower range
- Calculate y-intercept as percentage of response at 100% concentration:
- Intercept should be <= 2% of 100% response for assay methods
- Establish validated range as interval between lowest and highest concentrations for which linearity, accuracy, precision demonstrated
Got: Linear regression with R^2 >= 0.999 (assay) or >= 0.99 (impurity), random residual distribution, intercept <= 2% of target response, validated range clearly defined.
If fail: R^2 below criterion? Check preparation errors, detector non-linearity (too high concentration), analyte instability. Repeat with fresh preparations. Non-linearity inherent? Use polynomial calibration or narrow range.
Step 4: Assess Accuracy
- Prepare accuracy samples at 3 concentration levels (typically 80%, 100%, 120% of target for assay; LOQ, mid, high for impurity methods)
- At each level, prepare 3 independent replicates (minimum 9 determinations total)
- For drug substance — compare found concentration to known (gravimetric) amount
- For drug product — use spiked placebo approach: add known amounts of analyte to placebo matrix. Measure recovery
- Calculate percent recovery at each level:
- Recovery (%) = (found amount / added amount) x 100
- Acceptance criteria:
| Method Type | Recovery Range | RSD at Each Level |
|---|---|---|
| Assay (drug substance) | 98.0-102.0% | <= 2.0% |
| Assay (drug product) | 98.0-102.0% | <= 2.0% |
| Impurity (quantitation) | 80-120% at LOQ, 90-110% at higher levels | <= 10% at LOQ, <= 5% at higher |
| Cleaning validation | 70-130% (or tighter per company SOP) | <= 15% |
- Report individual recoveries, mean recovery, RSD at each level
Got: Mean recovery within acceptance criteria at all concentration levels, RSD within limits.
If fail: Recovery consistently high or low across all levels? Suspect systematic error in reference standard, sample preparation, or method (e.g., matrix effect causing ion suppression in LC-MS). Recovery varies erratic? Investigate sample preparation technique, analyte stability.
Step 5: Determine Precision
Evaluate three levels of precision:
- Repeatability (intra-day):
- One analyst, one instrument, one day
- Inject 6 determinations at 100% or 3 levels x 3 replicates (same data as accuracy)
- Calculate RSD of results: <= 2.0% for assay, <= 10% at LOQ for impurity
- Intermediate precision (inter-day / inter-analyst):
- Repeat repeatability study with different analyst, different day, and (if available) different instrument
- Calculate overall RSD combining both data sets
- Overall RSD <= 3.0% for assay
- Intermediate precision significantly worse than repeatability? Investigate source of variation (analyst technique, instrument calibration, environmental conditions)
- Reproducibility (for method transfer or multi-site validation):
- Performed at receiving laboratory following same protocol
- Compare results between laboratories
- Evaluated by F-test (variance comparison) and t-test (mean comparison) or equivalence testing
| Precision Level | Design | Acceptance (Assay) | Acceptance (Impurity Quant) |
|---|---|---|---|
| Repeatability | n >= 6 at 100%, 1 analyst, 1 day | RSD <= 2.0% | RSD <= 10% at LOQ, <= 5% above |
| Intermediate | 2 analysts, 2 days (or 2 instruments) | RSD <= 3.0% | RSD <= 15% at LOQ, <= 10% above |
| Reproducibility | Multi-laboratory | Per protocol / transfer criteria | Per protocol / transfer criteria |
Got: Repeatability and intermediate precision RSDs within acceptance criteria. No statistically significant difference between analysts/days/instruments beyond allowed RSD.
If fail: Intermediate precision much worse than repeatability? Identify variable driving additional variance (analyst preparation technique, ambient temperature, instrument calibration drift). Control before repeating.
Step 6: Establish LOD, LOQ, and Robustness
Limit of Detection (LOD) and Limit of Quantitation (LOQ):
- Calculate LOD and LOQ using signal-to-noise approach or standard deviation approach:
- LOD = 3.3 x (sigma / S) where sigma = standard deviation of response at low concentration, S = slope of calibration
- LOQ = 10 x (sigma / S)
- Alternative: S/N approach — LOD corresponds to S/N >= 3, LOQ to S/N >= 10
- Confirm experimental — prepare solutions at calculated LOD and LOQ concentrations and inject:
- At LOD: peak should be detectable but not necessarily quantifiable with acceptable precision
- At LOQ: inject 6 replicates and confirm RSD <= 10% and accuracy within 80-120%
- Report LOD and LOQ with method used for determination
Robustness:
- Identify critical method parameters (typically 5-7 factors):
- Mobile phase composition (+/- 2% organic)
- Mobile phase pH (+/- 0.2 units)
- Column temperature (+/- 5 C)
- Flow rate (+/- 10%)
- Detection wavelength (+/- 2 nm)
- Column lot/batch (if available)
- Vary each parameter deliberate within specified range while holding others constant (or use fractional factorial design for efficiency)
- Evaluate impact on system suitability parameters (retention time, resolution, tailing, area)
- Parameters that cause system suitability failure within tested range must be tightly controlled, documented as critical method parameters
- Summarize robustness results in table showing each varied parameter, range tested, impact on key responses
Got: LOD and LOQ experimentally confirmed. Robustness study completed with critical method parameters identified, control limits established.
If fail: LOQ precision exceeds 10% RSD? Method sensitivity insufficient at that concentration. Options: increase injection volume, concentrate sample, improve sample cleanup, or use more sensitive detector. Parameter shows method not robust (fails SST with small deliberate variation)? Tighten control of that parameter in method. Flag during method transfer.
Checks
- Method category identified, all required parameters determined per ICH Q2(R2)
- Validation protocol written with pre-defined acceptance criteria
- Specificity demonstrated — no interferences, degradation products resolved, peak purity confirmed
- Mass balance within 95-105% for forced degradation study
- Linearity established with R^2 >= 0.999 (assay) or >= 0.99 (impurity), residuals random
- Accuracy demonstrated at 3 levels with recovery within acceptance criteria
- Repeatability RSD within limits (e.g., <= 2.0% for assay)
- Intermediate precision RSD within limits (e.g., <= 3.0% for assay)
- LOD and LOQ experimentally confirmed (LOQ precision <= 10% RSD)
- Robustness study completed with critical method parameters identified
- All raw data, calculations, chromatograms compiled into validation report
Pitfalls
- Start experiments before protocol approval: In GMP environments, validation data generated before protocol approval may not be acceptable to regulators. Always obtain approval first.
- Use serial dilutions for linearity: Serial dilutions propagate pipetting errors. Prepare each concentration level independently from common stock for most accurate linearity assessment.
- Insufficient forced degradation: Too little degradation (< 5%) may miss important degradation products. Too much (> 30%) produces secondary degradation products that complicate interpretation. Target 5-20% degradation per condition.
- Confuse repeatability with intermediate precision: Repeatability = same-day, same-analyst, same-instrument. Intermediate precision must vary at least one of these factors. Both required for Category I and II methods.
- Neglect LOQ verification step: Calculating LOQ from calibration curve not sufficient. Calculated LOQ must be experimentally confirmed by demonstrating acceptable precision and accuracy at that concentration.
- Omit robustness until late in validation: Discovering method not robust after accuracy and precision studies wastes time and materials. Perform quick robustness screen early in validation to catch fragile parameters.
- Incomplete validation reports: Regulatory reviewers expect all raw data, chromatograms (not tabulated numbers), statistical analysis, explicit pass/fail conclusions for each parameter. Missing data leads to deficiency letters.
See Also
develop-gc-method-- GC method development that precedes validationdevelop-hplc-method-- HPLC method development that precedes validationinterpret-chromatogram-- reading chromatograms generated during validation experimentstroubleshoot-separation-- resolving issues discovered during validation studiesconduct-gxp-audit-- auditing completed validation for GxP compliancewrite-standard-operating-procedure-- documenting validated method as SOP
GitHub 仓库
相关推荐技能
evaluating-llms-harness
测试该Skill通过60+个学术基准测试(如MMLU、GSM8K等)评估大语言模型质量,适用于模型对比、学术研究及训练进度追踪。它支持HuggingFace、vLLM和API接口,被EleutherAI等行业领先机构广泛采用。开发者可通过简单命令行快速对模型进行多任务批量评估。
cloudflare-cron-triggers
测试这个Claude Skill提供了关于Cloudflare Cron Triggers的完整知识库,用于通过cron表达式定时执行Workers。它支持配置周期性任务、维护作业和自动化工作流,并能处理常见的cron触发错误。开发者可以用它来设置定时任务、测试cron处理器,并集成Workflows和Green Compute功能。
webapp-testing
测试该Skill为开发者提供了基于Playwright的本地Web应用测试工具集,支持自动化测试前端功能、调试UI行为、捕获屏幕截图和查看浏览器日志。它包含管理服务器生命周期的辅助脚本,可直接作为黑盒工具运行而无需阅读源码。适用于需要快速验证本地Web应用界面和交互功能的开发场景。
finishing-a-development-branch
测试这个Skill用于开发分支完成后的集成决策,当代码实现完成且测试通过时,它会引导开发者选择合适的工作流。它首先验证测试状态,然后提供合并、创建PR或清理等结构化选项。核心价值在于确保代码质量的同时,标准化分支收尾流程。
