05.27 class
Personal Genomics - Michael Cariaso
cost per raw megabase of DNA sequence keeps decreasing
23andMe customer growth
- buy one and spit on it and then they will send you back your genomic information (e.g. disease risk ,etc)
Since 2013 12(Dec), commercial advertisement started, ancestry research
pre-FDA problems 23andME = 107 things
-genetic risk factors (AD), inherited conditions (Bloom's syndrome, Canavan disease), drug response (abacavir hypersensitivity, acetaldehyde toxicity, clopidogrel plavix), traits (alcohol flush reaction, asparagus metabolite detection, bitter taste perception)
post-FDA problems 23andME = medical stuff gone
-genetic risk factors and drug response were all gone ....
KFDA Jun30, 2016 will soon announce the rules
23andME data
- rsid chromosome position genotype
reference SNP
if it's never been seen before, what does it mean?
sometimes we can make a good guess. synonymous = no effect frameshift = no effect 1 bad = carrier
loss of function SNPs with negligible effect
SNPedia (single nucleotide polymorphism wikipedia)
Promethease - a program that takes a file of genotypes with dbSNP (function like an enzyme that speeds up the reaction). you can know related disease onset rate, information about genetic magnitude with rsid, can control the amount of information shown (ex. heart-related disease, to show frequency <30% etc)
goal : make it possible for everyone to search about very personal genome by oneself and only to onself
boring genome = good and healthy person, interesting genome = bad
reality is majority do not want to know... why? (personally, I cannot understand)
Genomics - 이충연
introduction
Approaches to genetic association studies - 안수균
(1) candidate gene studies : set a hypothesis → target gene search. You can research specific target gene deeply, disadvantage : unpredictable genes are excluded. hypothesis-driven approach, relies on the current understanding of the biology, large scale & meta-analysis
(2) genome wide linkage studies :genome sequencing hypothesis-generating(data-driven) method
(3) genome wide association studies : hypothesis-generating method
limitation : geneome does not tell everything
hypothesis-driven approach and data-driven approach (you cannot say which one is better. need to pay close attention when to choose)
for genomics, data-driven approach is more appropriate
genomic T (우)
prediction is what we do - all algorithms in the world is to predict something
genome, envirome, phenome
Those are very complicated . challenging to expect percisely