All chapters
NGS Workflow Overview
beginnerNGS Pipeline at a Glance
Standard WES Pipeline Steps
🧪
Sample & Library Prep
DNA extraction → fragmentation → adapter ligation → capture (WES)
⚡
Sequencing
Illumina NovaSeq → FASTQ generation; target 100x mean depth
🔍
QC & Trimming
FastQC + fastp; check Q30, adapter content, GC bias
🗺
Alignment
BWA-MEM2 → hg38; sort + index BAM; flagstat QC
⚙️
BAM Processing
MarkDuplicates + BQSR; coverage metrics with mosdepth
🎯
Variant Calling
GATK HaplotypeCaller GVCF mode → joint genotyping → VQSR
📝
Annotation
ANNOVAR/VEP; gnomAD, ClinVar, CADD, ACMG criteria
📋
Clinical Report
ACMG classification + Exomiser phenotype prioritization
End-to-End Pipeline
Sample Collection
DNA Extraction
QC (NanoDrop/Qubit/Bioanalyzer)
Library Preparation
Capture (WES)
Sequencing
FASTQ Generation
Raw QC
Trimming
Alignment
BAM Processing
Variant Calling
Filtering
Annotation
Clinical Interpretation
Report
Pre-Sequencing (Wet Lab)
- DNA extraction: blood (EDTA tube), saliva, FFPE tissue, buccal swab
- DNA QC: NanoDrop (purity A260/A280 ≥1.8), Qubit (concentration ≥50 ng/µL), Bioanalyzer/TapeStation (integrity DIN/RIN ≥7)
- Library prep: fragmentation → end repair → A-tail → adapter ligation → PCR amplification
- WES capture: hybridisation with biotinylated baits, streptavidin bead pulldown, 2 rounds of capture
- Library QC: Bioanalyzer peak at ~300–400 bp, qPCR for quantification
- Pooling: normalise library concentrations, pool for multiplexed sequencing
Bioinformatics Steps
- 1. Raw QC - FastQC/MultiQC: check quality, adapter content, GC bias
- 2. Trimming - fastp/Trimmomatic: remove adapters, low-quality bases
- 3. Alignment - BWA-MEM2: map to hg38 reference genome
- 4. Sort & Index - samtools sort/index: required for downstream tools
- 5. Mark Duplicates - GATK MarkDuplicates: flag PCR duplicates
- 6. BQSR - GATK BaseRecalibrator + ApplyBQSR: correct systematic base quality errors
- 7. Variant Calling - GATK HaplotypeCaller: generate g.VCF per sample
- 8. Genotyping - GATK GenotypeGVCFs: joint calling across samples
- 9. Variant Filtration - VQSR or hard filters: remove low-quality calls
- 10. Annotation - ANNOVAR/VEP: add functional and clinical context
- 11. Prioritisation - filter by MAF, impact, HPO phenotype
Quality Checkpoints
- Mean coverage ≥100× for WES (clinical)
- Uniformity: ≥80% of target at ≥20×
- On-target rate ≥60%
- Duplicate rate <20%
- Ts/Tv ratio: ~2.0–2.1 for WGS, ~3.0–3.3 for WES (high ratios = good calls)
- Het/Hom ratio: ~1.5–2.5 for germline (deviation suggests contamination or errors)
- Contamination: <3% (VerifyBamID FREEMIX score)