Module 5: Genomics and Functional Genomics
Functional Genomics
Definition: Characterizes the function of sequences (mRNA, protein).
Predicting Gene Functions
Bioinformatics: Predicts function from sequence alone.
Homology Searching: Identifies orthologs in other species to infer function; requires phylogenetic analysis.
Protein Domains: Regions within proteins with specific shapes/functions; scanning databases infers function.
Experimental Mutagenesis: Random generation of mutations (radiation, chemicals, transposable elements, CRISPR-Cas) to produce observable phenotypes for screening.
Genome-Wide Mutagenesis
Relies on random mutation generation to produce observable phenotypes.
Offspring screened for mutant phenotypes; dominant mutations in heterozygotes, recessive require further crosses.
Positional cloning identifies mutated gene after phenotype detection.
The ENCODE Project
Aim: Identify all functional elements (transcription, chromatin, splicing) in human and mouse genomes.
Medical Value: Invaluable for inferring function of non-coding regions where most disease mutations occur.
Medical Genomics and GWAS
SNPs (Single Nucleotide Polymorphisms): Single base pair differences, inherited, usually benign, c. rac{1}{1000} bp.
SNP Haplotypes: Physically linked SNPs inherited together, can mark disease-causing loci.
Genome-Wide Association Studies (GWAS): Compares genetic variations (SNPs) between disease and control groups to find associations, pointing to disease-causing genomic regions.
Calculating Odds Ratio (GWAS)
Formula: OR = \frac{\text{Odds of exposure in cases}}{\text{Odds of exposure in controls}} = \frac{a \times d}{b \times c}.
a: cases with variant; b: controls with variant.
c: cases without variant; d: controls without variant.
Interpretation:
OR = 1: No association between variant and disease.
OR > 1: Variant increases disease risk.
OR < 1: Variant decreases disease risk.
Genomic Variation Databases
ClinVar: Genomic variation and human health.
MedGene: Human medical genetics conditions.
OMIM: Human genes and genetic phenotypes.
GTR: Genetic test information.
dbSNP: SNPs and small indels.
dbVar: Structural variants.
Personalized Genomics (e.g., 23andMe)
Direct-to-consumer genetic testing for health and ancestry.
Initially faced FDA restrictions on health reports, later approved for specific disease risk and carrier status tests.
Pharmacogenetics
Genetic testing to predict individual drug response.
Warfarin Case Study: Variations in VKORC1 and CYP2C9 affect metabolism/sensitivity, but genetic testing has poor predictive value for dosing in large populations; clinical guidelines advise against routine use.
Limitations of Genomics for Complex Diseases
Low Predictive Value: GWAS often identifies genes that explain only a small percentage of total genetic variation for complex traits/diseases.
Multigenic diseases require more sophisticated approaches, e.g., Machine Learning methods like Random Forest.