Genomics
Learning Objectives
- Master eukaryotic chromosome compaction hierarchies from the 2 nm Watson-Crick double helix to 1400 nm metaphase structures.
- Understand the functional organization of non-coding regulatory elements: promoters, enhancers, silencers, and CTCF insulators.
- Characterize the human genetic variation spectrum: Single Nucleotide Variants (SNVs), small indels, Copy Number Alterations (CNAs), and balanced Structural Variants (SVs).
- Execute a standard production germline variant calling pipeline using BWA-MEM2 and GATK4 HaplotypeCaller.
1. The Physical Architecture of the Eukaryotic Genome
Genomics is the study of the complete genetic blueprint of an organism, encompassing its primary DNA sequence, epigenetic modifications, higher-order three-dimensional nuclear chromatin architecture, and evolutionary conservation. The human diploid genome contains approximately 6.2 billion base pairs distributed across 23 chromosome pairs. If completely unraveled, the linear DNA within a single human diploid nucleus would stretch approximately 2 meters in length, yet it is compacted into a nuclear envelope measuring just 5 to 10 micrometers in diameter.
This extraordinary compaction is achieved through a multi-tiered hierarchy of structural organization:
- The Double Helix (2 nm): Antiparallel sugar-phosphate backbones held together by Watson-Crick complementary hydrogen bonding (A-T with 2 hydrogen bonds, G-C with 3 hydrogen bonds). B-form DNA exhibits approximately 10.5 base pairs per helical turn with a 3.4 nm helical pitch.
- Nucleosomal Beads-on-a-String (11 nm): 147 base pairs of DNA are wound 1.65 times around an octameric protein core composed of two copies each of histones H2A, H2B, H3, and H4. Linker DNA (20–80 bp) bound by linker histone H1 connects adjacent nucleosomes.
- Chromatin Fiber & Loop Domains (30–100 nm): Nucleosomes condense into higher-order chromatin fibers organized into Topologically Associating Domains (TADs), stabilized by the ring-shaped cohesin complex and CCCTC-binding factor (CTCF) insulators.
- Metaphase Chromosome (1400 nm): During mitosis, chromatin reaches maximal condensation, forming sister chromatids joined at the centromere and capped at terminal ends by protective telomeric repeat hexamers (TTAGGG)n.
2. Non-Coding Architecture & The Transposon Landscape
Remarkably, protein-coding exons comprise merely 1.2% to 1.5% of the human genome (~20,000 protein-coding genes). The remaining >98% consists of non-coding elements that govern genome architecture and spatial gene regulation:
- Promoters: Core sequences immediately upstream of the transcription start site (TSS), containing TATA boxes and initiator motifs bound by General Transcription Factors (GTFs) and RNA Polymerase II.
- Enhancers: Distal regulatory DNA elements (located up to 1 megabase away) containing dense clusters of transcription factor binding sites (TFBS) that physically loop to promoters via the Mediator complex.
- Silencers and Insulators: Insulators bound by CTCF establish chromatin loop boundaries, preventing inappropriate cross-talk between adjacent genomic neighborhoods.
- Transposable Elements (TEs): Account for nearly 45% to 50% of human DNA. Divided into:
- LINEs (Long Interspersed Nuclear Elements): Autonomous retrotransposons (e.g. LINE-1, ~6 kb), encoding reverse transcriptase and endonuclease.
- SINEs (Short Interspersed Nuclear Elements): Non-autonomous elements (e.g. Alu elements, ~300 bp), hijacking LINE-1 machinery for mobilization.
- LTR Retrotransposons & HERVs: Human Endogenous Retroviruses remnants of ancestral retroviral germline integrations.
3. The Genomic Variation Spectrum
Between any two unrelated human genomes, there is approximately 0.1% sequence divergence (representing ~4 to 5 million variants). These variants span several distinct classes:
| Variant Class | Size Scale | Genome Frequency (per person) | Detection Methodology |
|---|---|---|---|
| Single Nucleotide Variants (SNVs) | 1 base pair | 3,500,000 - 4,000,000 | Short-read NGS, GATK HaplotypeCaller, DeepVariant |
| Small Insertions / Deletions (Indels) | 1 - 49 base pairs | 500,000 - 800,000 | Local de novo re-assembly (GATK, Strelka2) |
| Structural Variants (SVs) | ≥ 50 bp to megabases | 20,000 - 25,000 | Long reads (Sniffles2, hifiasm), Split-read short (Manta) |
| Copy Number Alterations (CNVs) | 1 kb to whole chromosome | 1,000 - 2,500 | Read depth analysis (CNVkit, Control-FREEC), Arrays |
4. Production Germline Analysis Pipeline
The standard GATK Best Practices pipeline for mapping short reads, marking PCR duplicates, and discovering germline variants:
# 1. Align trimmed paired-end reads to reference genome (GRCh38)
bwa-mem2 mem -t 16 -R "@RG\tID:sample1\tSM:sample1\tPL:ILLUMINA" \
GRCh38.primary_assembly.fa sample_R1.fastq.gz sample_R2.fastq.gz | \
samtools sort -@ 8 -m 2G -o sample_aligned.bam -
# 2. Mark optical and PCR duplicate fragments
gatk MarkDuplicatesSpark \
-I sample_aligned.bam \
-O sample_dedup.bam \
-M duplicate_metrics.txt
# 3. Discover germline SNPs and indels via local de Bruijn assembly
gatk HaplotypeCaller \
-R GRCh38.primary_assembly.fa \
-I sample_dedup.bam \
-O sample_variants.vcf.gz \
-ERC GVCF
# 4. Joint genotyping and VCF quality filtering
gatk GenotypeGVCFs \
-R GRCh38.primary_assembly.fa \
-V sample_variants.vcf.gz \
-O sample_genotyped.vcf.gz