Types of Sequencing

Selecting the correct sequencing strategy is one of the most impactful decisions in a genomics study. Each approach has distinct trade-offs in cost, coverage, and the types of variants it can detect. A common mistake is ordering WGS when a targeted panel would be cheaper, faster, and equally informative - or conversely, ordering a panel when the patient's phenotype is not explained by known genes.

• WGS (Whole Genome Sequencing): sequences all 3.2 Gb; 30× depth ≈ $200–400; detects SNVs, indels, SVs, CNVs, non-coding variants; gold standard for research; increasingly used clinically for unsolved rare disease
• WES (Whole Exome Sequencing): ~50 Mb coding target (1.5% of genome); 100× depth ≈ $100–250; detects ~95% of known Mendelian disease variants; clinical standard for rare disease
• Targeted gene panel: 5–500 genes; 500–2000× depth; cheapest per test ($100–500); fastest turnaround; used in hereditary cancer, cardiology, pharmacogenomics, oncology
• RNA-seq (whole transcriptome): measures gene expression levels and isoforms; detects gene fusions (cancer); no stable DNA required; captures splicing defects WES misses
• Bisulfite sequencing (WGBS, RRBS): maps 5-methylcytosine genome-wide; epigenetic studies; imprinting disorders (Prader-Willi, Angelman)
• ChIP-seq: maps protein-DNA interactions; transcription factor binding sites; histone modifications; requires antibody for target protein
• ATAC-seq: open chromatin profiling; identifies active regulatory elements; single-cell ATAC-seq (scATAC-seq) maps cell-type specific accessibility

WES captures ~180,000 exons across ~20,000 genes. The "exome" includes coding exons ± 20 bp of flanking intronic sequence to catch splice-site variants. Three capture kit families dominate clinical practice, each with different gene coverage and target sizes.

• Agilent SureSelect V8: ~35.8 Mb target; ~20,000 genes including UTRs; most widely used in India and Asia
• Illumina Nextera Flex for Enrichment: ~45 Mb; includes UTRs and miRNA; optimised for NovaSeq
• Twist Bioscience Core Exome: ~33 Mb; very high uniformity; preferred for clinical labs needing consistent coverage
• Coverage metric - mean depth: 100× clinical, 50× research minimum; each additional 10× costs ~10% more
• Coverage metric - uniformity: target ≥80% bases at ≥20×; poor uniformity = GC-biased library prep or poor DNA quality
• Coverage metric - on-target rate: typically 60–80% of reads align to target; lower = excessive off-target amplification
• Sensitivity: WES at 100× detects >99% of heterozygous SNVs in target regions; ~80% sensitivity for small indels ≤10 bp
• Known blind spots: pseudogenes (PMS2, CYP21A2), segmental duplications (SMN1/SMN2), GC-extreme regions (KCNQ1)

Coverage is not just about the mean depth - it's about where your reads are NOT. A sample with mean 120× but 15% of target below 20× will miss variants in those regions. Always examine the coverage distribution, not just the mean.

• Coverage formula: depth = (read count × read length) ÷ target size; e.g., 50M paired reads × 150 bp ÷ 50 Mb = 150× WES
• 30× WGS: germline SNV sensitivity ~99%; adequate for population studies, most SVs; misses low-VAF somatic variants
• 100× WES: clinical germline standard; detects heterozygous variants (VAF ~50%) reliably at DP ≥30
• 200-500× targeted: somatic variant detection down to VAF 2-5%; required for liquid biopsy tumour panels
• 1000-2000× targeted: detects VAF <1%; used for minimal residual disease (MRD) monitoring in AML
• UMI (Unique Molecular Identifier) deduplication: at >200× depth, must use UMIs to distinguish true rare variants from PCR errors
• GC bias: low GC (<25%) and high GC (>70%) regions have lower coverage; BRCA1 exon 1 (GC ~80%) is often undercovered
• Rule: if a clinical WES report notes ≥10 genes with <20× coverage in relevant regions, request re-sequencing or targeted follow-up

A 35-year-old woman diagnosed with breast cancer at age 32 is referred for genetic counselling. Her maternal aunt had ovarian cancer. The ordering clinician must choose between a hereditary cancer gene panel (HBOC panel, ~25 genes) and clinical WES.

• HBOC panel chosen: covers BRCA1, BRCA2, PALB2, ATM, CHEK2, RAD51C, RAD51D, BRIP1 at very high depth (1000×)
• Result: BRCA2 c.5946delT p.(Ser1982Argfs*22) - Pathogenic frameshift; detected at 48% VAF (expected ~50%)
• Why not WES? Panel is cheaper ($500 vs $1500), faster (1 week vs 3 weeks), and provides deeper coverage of target genes
• WES would be chosen if: family has unusual features (male breast cancer + neurological symptoms → consider PTEN), no finding on panel, or phenotype not explained by known genes
• Clinical impact: patient eligible for olaparib (PARP inhibitor); bilateral risk-reducing mastectomy counselled; family cascade testing with BRCA2 single-site test ($200)