interpret-chromatogram
について
このスキルは、システム適性の検証、ピークの同定と積分、クロマトグラフィック特性値の計算を通じて、GC/HPLCクロマトグラムを解析します。保持時間とスペクトルマッチングによるピーク品質の評価により、信頼性の高い定性・定量分析を実現します。開発者はこれを用いて、正確な結果を得るためのクロマトグラフィーデータ処理を自動化できます。
クイックインストール
Claude Code
推奨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/interpret-chromatogramこのコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします
ドキュメント
クロマトグラムの解釈
Systematic interpretation of GC and HPLC chromatograms covering system suitability verification, peak identification, integration, calculation of chromatographic parameters, and assessment of peak quality for confident qualitative and quantitative results.
使用タイミング
- Reviewing chromatographic data before reporting analytical results
- Verifying that a system suitability test passes before running a sample sequence
- Identifying unknown peaks or confirming known analytes by retention time or spectral data
- Troubleshooting unexpected peaks, baseline anomalies, or integration artifacts
- Training analysts on chromatographic data interpretation
入力
Required
- Chromatogram data: Digital or printed chromatogram with time axis and detector response axis
- Reference standard data: Retention times and responses of known analytes under the same method conditions
- Method parameters: Column, mobile phase/carrier gas, temperature/gradient program, detector settings
Optional
- Spectral data: UV-Vis spectra (DAD), mass spectra (MS), or other spectral information for peak confirmation
- Previous chromatograms: Historical data from the same method for trend comparison
- System suitability criteria: Acceptance limits from the method or regulatory standard
- Sample preparation details: Dilution factors, extraction recovery, internal standard concentration
手順
ステップ1: Verify System Suitability
Confirm that the chromatographic system is performing within specification before interpreting sample data.
| Parameter | Typical Specification | Calculation |
|---|---|---|
| Retention time RSD | <= 1.0% | RSD of tR over n >= 5 injections |
| Peak area RSD | <= 2.0% (assay), <= 5.0% (impurity) | RSD of area over n >= 5 injections |
| Tailing factor (T) | 0.8-2.0 (USP), ideally 0.9-1.2 | T = W0.05 / (2 * f) where W0.05 = width at 5% height, f = front half-width |
| Resolution (Rs) | >= 1.5 (baseline), >= 2.0 (regulated) | Rs = 2(tR2 - tR1) / (w1 + w2) |
| Theoretical plates (N) | Per column spec (e.g., >= 2000) | N = 16(tR / w)^2 or N = 5.54(tR / w0.5)^2 |
| Capacity factor (k') | 2.0-10.0 for primary analyte | k' = (tR - t0) / t0 |
- Locate the system suitability injections (typically 5-6 replicates of a reference standard at the start of the sequence).
- Calculate each parameter from the table above.
- Compare calculated values against the method's acceptance criteria.
- If any parameter fails, the system is not suitable -- do not proceed to sample interpretation until the issue is resolved.
- Document all system suitability results in the batch record.
期待結果: All system suitability parameters within specification, confirming the system is fit for purpose.
失敗時: If retention time RSD fails, check for temperature instability, mobile phase preparation errors, or column degradation. If tailing factor fails, inspect the inlet liner (GC) or column frit (HPLC). If resolution fails, check column performance with a dedicated test mix and replace if necessary.
ステップ2: Identify Peaks
- Compare each peak's retention time (tR) against the reference standard chromatogram.
- Acceptable retention time match: within +/- 2% of the reference tR (or +/- 0.1 min for short runs).
- For ambiguous identifications, use co-injection (spiking): add reference standard to the sample and re-inject. The target peak should increase without broadening or shouldering.
- For DAD-equipped HPLC: compare the UV-Vis spectrum of each peak against a spectral library.
- Spectral match index >= 990 (out of 1000) for positive identification.
- Check spectral purity across the peak (front, apex, tail spectra should overlay).
- For MS-equipped systems: confirm molecular ion (m/z) and key fragment ions against reference spectra.
- Flag any peak that cannot be identified -- report it as "unknown" with its retention time and relative response.
期待結果: All target analytes identified by retention time matching, with spectral confirmation where available. Unknown peaks flagged with retention time and area.
失敗時: If retention times have shifted uniformly, a systematic change has occurred (column aging, temperature drift, mobile phase error). Re-inject the reference standard to establish current retention times before re-evaluating.
ステップ3: Perform Peak Integration
- Select integration mode:
- Automatic integration with data system defaults as a starting point
- Manual adjustment only when automatic integration demonstrably misplaces baseline or peak boundaries
- Set integration parameters:
- Baseline detection sensitivity (slope sensitivity / threshold)
- Minimum peak area or height to reject noise
- Peak width parameter matching the narrowest expected peak
- Verify baseline placement:
- Baseline should connect the start and end of each peak at the true chromatographic baseline
- For overlapping peaks, use valley-to-valley or perpendicular drop methods as specified by the method
- For gradient methods, baseline may rise -- use a tangent skim or exponential skim for peaks on a rising baseline
- Check for integration errors:
- Split peaks integrated as two when they should be one
- Shoulder peaks merged into the main peak when they should be separate
- Noise spikes integrated as peaks
- Baseline drawn through a peak (negative peak clipping)
- Record the final integration parameters and any manual adjustments with justification in an audit trail.
期待結果: All target peaks integrated with correct baseline placement, no artifacts included, and all manual adjustments documented with rationale.
失敗時: If the automatic integrator consistently mishandles a particular peak shape, create a timed-events integration method with custom parameters for that retention window. Never manually adjust integration to achieve a desired result -- adjustments must be scientifically justified.
ステップ4: Calculate Chromatographic Parameters
Calculate the following for all reported peaks:
- Resolution (Rs) between adjacent peaks:
- Rs = 2(tR2 - tR1) / (w1 + w2)
- Rs >= 1.5 indicates baseline separation; Rs >= 2.0 provides margin for routine use
- Tailing factor (T) at 5% peak height:
- T = W0.05 / (2f)
- T = 1.0 is perfectly symmetric; T > 2.0 indicates significant tailing
- Theoretical plates (N):
- N = 16(tR / w)^2 using baseline width, or N = 5.54(tR / w0.5)^2 using half-height width
- Higher N means better column efficiency
- Capacity factor (k'):
- k' = (tR - t0) / t0 where t0 is the dead time (void volume / flow rate)
- Ideal range: 2-10 for good separation with reasonable run time
- Selectivity factor (alpha) between critical pair:
- alpha = k'2 / k'1
- alpha > 1.05 is generally needed for adequate separation
- Tabulate results for all analytes and compare against method specifications.
期待結果: All chromatographic parameters calculated, tabulated, and compared to acceptance criteria. Critical pair resolution and plate count documented.
失敗時: If calculated plates are significantly below the column specification, the column may be degraded -- test with a fresh standard and compare to historical data. If parameters drift within a sequence, investigate instrument stability.
ステップ5: Assess Peak Quality
- Symmetry: Peaks should be Gaussian or near-Gaussian. Significant fronting (T < 0.8) suggests column overload; tailing (T > 1.5) suggests secondary interactions or dead volume.
- Baseline separation: For quantitative work, critical pairs must be baseline-resolved. If valley between peaks does not return to baseline, note the percentage valley and assess impact on accuracy.
- Peak width consistency: Peaks that are significantly broader than expected (compared to the standard) may indicate on-column degradation, extra-column band broadening, or injection issues.
- Spectral purity (DAD/MS): If the purity index indicates spectral inhomogeneity across the peak, a co-eluting impurity is likely present. Consider adjusting the method for better resolution.
- Negative peaks or baseline disturbances: Negative peaks in UV indicate the sample solvent absorbs more than the mobile phase at the detection wavelength -- this is normal for the solvent front but abnormal elsewhere.
- Ghost peaks: Peaks present in the blank injection indicate carryover, contaminated mobile phase, or column bleed. Identify the source before reporting sample results.
- Summarize overall chromatographic quality and note any limitations on the reported results.
期待結果: Peak quality assessed for all target analytes; any anomalies (tailing, co-elution, ghost peaks) documented with their potential impact on data quality.
失敗時: If significant quality issues are found (co-elution confirmed by spectral impurity, ghost peaks at analyte retention times), the data may not be reportable. Flag the results, investigate root cause, and re-run after corrective action.
バリデーション
- System suitability parameters calculated and within specification
- All target analytes identified by retention time (+/- spectral confirmation)
- Unknown peaks flagged with retention time and area
- Integration performed with correct baseline placement; manual adjustments documented
- Resolution, tailing, plates, and capacity factor calculated for all peaks
- Peak quality assessed -- no unresolved co-elutions affecting quantitation
- Ghost peaks and carryover evaluated via blank injection
- Results tabulated and compared against method acceptance criteria
よくある落とし穴
- Accepting automatic integration without review: Data systems can misplace baselines, especially for shoulders, small peaks near large ones, and gradient baselines. Every chromatogram must be visually reviewed.
- Confusing retention time shift with a new peak: Uniform retention time shifts (all peaks move together) indicate a systematic change, not new compounds. Re-inject the standard to recalibrate before making identification calls.
- Reporting peaks below the noise level: Peaks with signal-to-noise ratio below 3 (detection) or 10 (quantitation) should not be identified or quantitated. Calculate S/N explicitly for trace-level peaks.
- Ignoring the solvent front: The void volume peak is not an analyte. Ensure t0 is correctly identified and excluded from analyte reporting.
- Manual integration to achieve a target result: Adjusting integration to make a result pass specification is data falsification. All integration changes must be scientifically justified and audit-trailed.
- Neglecting spectral purity checks: A clean-looking peak can hide a co-eluting impurity. Always check peak purity when DAD or MS data is available.
関連スキル
develop-gc-method-- method development for the GC technique producing the chromatogramdevelop-hplc-method-- method development for the HPLC technique producing the chromatogramtroubleshoot-separation-- diagnosing problems identified during chromatogram interpretationvalidate-analytical-method-- formal validation of the method generating the chromatographic datainterpret-mass-spectrum-- detailed interpretation of MS data for GC-MS and LC-MS peak confirmation
GitHub リポジトリ
関連スキル
executing-plans
デザインexecuting-plansスキルは、完全な実装計画があり、それを管理されたバッチでレビューチェックポイントを設けながら実行する場合に使用します。このスキルは計画を読み込んで批判的にレビューした後、小さなバッチ(デフォルトは3タスク)でタスクを実行し、各バッチの間に進捗状況を報告してアーキテクトのレビューを受けます。これにより、品質管理チェックポイントが組み込まれた体系的な実装が保証されます。
requesting-code-review
デザインこのスキルは、コードレビュアーサブエージェントを起動し、処理を進める前に要件に対してコード変更を分析します。タスク完了後、主要な機能の実装後、またはmainブランチへのマージ前などに使用すべきです。このレビューは、現在の実装と元の計画を比較することで、問題を早期に発見するのに役立ちます。
connect-mcp-server
デザインこのスキルは、開発者がHTTP、stdio、またはSSEトランスポートを使用してMCPサーバーをClaude Codeに接続するための包括的なガイドを提供します。GitHub、Notion、カスタムAPIなどの外部サービスを統合するためのインストール、設定、認証、セキュリティについて解説しています。MCP統合のセットアップ、外部ツールの設定、またはClaudeのModel Context Protocolを扱う際にご利用ください。
web-cli-teleport
デザインこのスキルは、タスク分析に基づいて開発者がClaude Code WebとCLIインターフェースの選択を支援し、これらの環境間でのシームレスなセッションテレポーテーションを可能にします。Web、CLI、モバイル環境を切り替える際のセッション状態とコンテキストを管理することで、ワークフローを最適化します。様々な段階で異なるツールを必要とする複雑なプロジェクトにご活用ください。
