McedClassifierQcChecker

URS & FS — User Requirements and Functional Specification

This document describes the controlled interface and behaviour of McedClassifierQcChecker for MCED Classifier QC Checker.

Domain limits and critical parameters

• Comprehensive verification of MCED classifier prediction reliability per FDA SaMD guidance, EU IVDR and MCED consortium standards: input data quality (cfDNA ≥10ng, CpG coverage ≥30×, bisulfite conversion ≥99%, fragment profile, background noise ≤0.15), feature extraction (completeness ≥90%, informative fragments ≥5000), classifier output (tissue confidence ≥0.70 for positives, model agreement ≥0.80), dual control verification (positive + negative), and technical flag detection (overtrimming, art
• ⚠️ Critical: MCED applies to ASYMPTOMATIC population — cost of error is extremely high. Incomplete bisulfite conversion mimics hypermethylation = false positive. Low model agreement = unstable prediction. Any control failure = TECHNICAL_FAIL. Positive result requires diagnostic confirmation; negative result does not rule out cancer.

URS — User Requirements Specification

input.csv contract

SampleID,ClassifierName,ClassifierVersion,CfDNA_Input_ng,MinCfDNA_Input_ng,CpG_Coverage_Mean_X,MinCpG_Coverage_X,BisulfiteConversion_Rate,MinBisulfiteConversion_Rate,FragmentSize_Peak_bp,FragmentProfile_Valid,BackgroundNoise_Score,MaxBackgroundNoise_Score,FeatureCompleteness_Percent,MinFeatureCompleteness_Percent,InformativeFragments_Count,MinInformativeFragments,CancerSignalScore,CancerSignalThreshold,PredictedTissue,TissueConfidence,MinTissueConfidence,ModelAgreement_Score,MinModelAgreement_Score,PositiveControl_Pass,NegativeControl_Pass,PosControl_Score_Observed,PosControl_Score_Expected_Min,NegControl_Score_Observed,NegControl_Score_Expected_Max,ArtifactDetected,OvertrimmingDetected,HemolysisIndex,MaxHemolysisIndex
MCED-2026-SCR-001,OncoDetect_MC,v3.2,25.0,10,48.0,30,0.996,0.99,167,true,0.06,0.15,97.5,90,22000,5000,0.92,0.5,Lung,0.88,0.70,0.94,0.80,true,true,0.91,0.8,0.03,0.1,false,false,8,50
MCED-2026-SCR-002,OncoDetect_MC,v3.2,18.0,10,22.0,30,0.972,0.99,145,false,0.28,0.15,72.0,90,3200,5000,0.45,0.5,Colon,0.52,0.70,0.55,0.80,false,true,0.42,0.8,0.05,0.1,true,true,65,50
MCED-2026-SCR-003,OncoDetect_MC,v3.2,30.0,10,55.0,30,0.998,0.99,168,true,0.04,0.15,98.2,90,28000,5000,0.03,0.5,None,0.0,0.70,0.97,0.80,true,true,0.89,0.8,0.02,0.1,false,false,5,50

Input validation rules

FS — Functional Specification

Example output.json

{
  "utilityId": "mcedclassifierqcchecker",
  "utilityFolder": "McedClassifierQcChecker",
  "package": "LiquidBiopsy",
  "overallStatus": "PASS|WARNING|FAIL",
  "sourceFile": "input.csv",
  "processedAtUtc": "2026-06-10T00:00:00Z",
  "checks": [
    {
      "parameter": "SampleID",
      "value": "MCED-2026-SCR-001",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-001"
    },
    {
      "parameter": "ClassifierName",
      "value": "OncoDetect_MC",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-002"
    },
    {
      "parameter": "ClassifierVersion",
      "value": "v3.2",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-003"
    },
    {
      "parameter": "CfDNA_Input_ng",
      "value": "25.0",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-004"
    },
    {
      "parameter": "MinCfDNA_Input_ng",
      "value": "10",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-005"
    },
    {
      "parameter": "CpG_Coverage_Mean_X",
      "value": "48.0",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-006"
    },
    {
      "parameter": "MinCpG_Coverage_X",
      "value": "30",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-007"
    },
    {
      "parameter": "BisulfiteConversion_Rate",
      "value": "0.996",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-008"
    },
    {
      "parameter": "MinBisulfiteConversion_Rate",
      "value": "0.99",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-009"
    },
    {
      "parameter": "FragmentSize_Peak_bp",
      "value": "167",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-010"
    },
    {
      "parameter": "FragmentProfile_Valid",
      "value": "true",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-011"
    },
    {
      "parameter": "BackgroundNoise_Score",
      "value": "0.06",
      "status": "PASS|WARNING|FAIL",
      "message": "Deterministic rule-based check result",
      "ruleReference": "FS-RULE-012"
    }
  ],
  "criticalFindings": [],
  "warnings": [],
  "audit": {
    "inputHash": "sha256:<calculated at runtime>",
    "rulesVersion": "<utility executable version>",
    "documentation": "McedClassifierQcChecker.documentation.html"
  }
}

Traceability matrix

IQ/OQ/PQ test scenarios

QA/QC and change control

• Do not rename columns without updating validator, documentation and test set.
• Retain input.csv, output.json, executable version and checksum.
• Before production use, perform IQ/OQ/PQ or equivalent CSV/CSA verification.
• Critical limits shall be verified against the approved specification, registration dossier and local SOPs.
• The utility provides structured QC decision support; final release decision remains with QA/QP and approved procedures.