1. Samples: tumour and matched normal
Why a normal sample is sequenced too, and how purity and FFPE change the data.
Whole-exome sequencing (WES)
From a tumour and a blood sample to a mutation landscape. Data: TCGA-LAML (shipped with maftools, 193 patients)
Bench & sequencing
2. Library prep and exome capture
Fragmentation, adapters, hybrid capture: how exons are fished out, and why the capture size matters for TMB.
3. Sequencing and coverage depth
Paired-end reads, depth, duplicates and uniformity, and how depth limits the lowest VAF you can detect.
From reads to data
4. Alignment: putting reads back on the genome
BWA alignment, BAM files, mapping quality, duplicate marking and base-quality recalibration.
5. Somatic variant calling
Tumour versus normal: which differences are somatic mutations, and which are germline variants or artefacts.
6. Annotation and the MAF file
Which gene a mutation hits and what it does to the protein; what each MAF column says.
Analysis
7. The cohort landscape: oncoplots
Read an oncoplot on TCGA-LAML: frequency, exclusivity and co-occurrence.
8. Tumour mutational burden (TMB)
TMB's numerator and denominator, where the 10 mut/Mb cut-off comes from and why it does not transfer freely.
9. Mutation spectra and signatures
Six substitution classes, 96 trinucleotide contexts, and how signatures point to mutational processes.
10. Driver genes and hotspots
Hotspots on a lollipop plot: what positional clustering finds, and what it misses.
11. Mutations and outcome
Survival of mutant versus wild-type, and how to define wild-type correctly.
12. Beyond: heterogeneity, cohort comparison, and OmicOne
VAF and clonal structure, comparing two cohorts, clinical enrichment, and the full run in OmicOne.