W8L2 Fri populational scale genomic

Question 1

gnomAD and ExAC

Answer 1

• Exome Aggregation Consortium (2016), ~60,000 mostly healthy human exomes.
• Followed by Genome Aggregation Consortium (2020), >125,748 exomes
  + 14,000 genomes (currently >75,000 genomes)
  -use to let us know if something is rare

Question 2

problem with understanding rare mutation and solution

Answer 2

• Knowledge of mutation pathogenicity is limited by small sample sizes.
• All of us carry rare variants.
• Differentiating the rare variants that cause Mendelian disease from the ones that are neutral is hard, precisely because they are all rare.
  -use exac/ gnomad to find which rare variant is benign/ disease associated

Question 3

An example: Rare prion disease in ExAC

Answer 3

• Sonia’s mother died in 2011 from fatal familial insomnia, a rare prion disease caused by mutations in the gene PRNP
• On the basis of ExAC data (possible have 2 case) but found 52 alleles in the Exac , the Vallabh-Minikels were able to confirm that Sonia is a carrier of a pathogenic allele and will develop FFI
  -compare gene in ExAC and case reported by prion surveillance center

Question 4

Are 150,000 samples enough

Answer 4

• ExAC and gnomAD give us a view of the mutational load in humans far more precise than anything before but this does not account for all possible mutation
  -still alot of no-CPG transition to be discover, CPG, we probably have seen alot

Question 5

Protein truncating variant in human

Answer 5

• we have 19% rare heterozygous and 1% homozygous PTV mutation per individual
  -some class of gene can be disrupted normally while other cant be

Question 6

Selective constraint and the human mutational load

Answer 6

• With enough data we can quantify selective constraint acting on genes. finding which gene is important and which is not
• pLoF: summarises the excess/deficit of loss of function mutations per gene
  -our gene isnt as resistance agaisnt PTV as we expected
  -look at how different gene tolerantcy to conclude their importance
  -need large data set

Question 7

Large cohorts are extremely valuable: example

Answer 7

• Myriad patented the sequence of BRCA1 in 1994 (1995 in Australia).
  Their genetic test (USD 2,000) was the only way women could lawfully access their BRCA1 sequence
• value of human each genome cost 75 -300 $ each

Question 8

UKBiobank

Answer 8

• ~500,000 individuals (UK individuals aged 50-70), 4000+ phenotypes from each of them.
• But only genotyped at ~1 million loci
• Cohort will be re-phenotyped multiple times in the future to gather data about late onset disease.
• Explicitly medical goals.
• 94% of participants self-report as ‘white’; 88% as ‘British’

Question 9

Some problem with Biobank ethical

Answer 9

• Poor definition of same-sex attraction (“ever having had sex with someone of the same sex”)
• Conflation with ‘risk-taking behaviour’ (homosexuality was illegal in the UK until recently)
• Unclear consent – UK Biobank is a medical resource.
• No engagement with LGBTQI+ community until first results were publicly presented.

Question 10

Predictive power of PRS in different ethic group

Answer 10

• PRS for height in data from GIANT(mostly pan- European samples)
• Performs worse elsewhere, relative
  to predictive accuracy in European populations
• Polygenic scores work a lot better in individuals from the same population as the one where the data come from