2-5-25 Flashcards
PGx drug discovery I (26 cards)
how does a 2x2 contingency table form from a case control study?
convert the number of gentoypes into number of alleles
what does a p value of >0.1 mean?
no presumption against the null hypothesis
no significant association
what does 0.05<P<0.1 mean?
low presumption against the null hypothesis
marginal association
what does P<0.05 mean?
strong presumption against the null hypothesis
significant association
what does P<0.01 mean?
very strong presumption against null hypothesis
very significant association
what does P value not meaure?
the strength of an association relationship
what can affect p value?
sample size –> the bigger the size, the lower the p value even under the same frequency
allele frequency
what is an odds ratio (OR)?
measure of strength
increased risk for a phenotype by carrying a specific genotype/allele compared to the patients without carrying
what does the OR measure in a simple terms?
odds of phenotype in an individual with the genotype/allele
odds of phenotype in an individual WO the genotype/allele
what is the hazard ratio (HR)?
similar concept to odds ratio, but mainly discussed in survival data
what does an OR of 1 mean?
no association
what does an OR of greater than 1?
potentially increases the risk –> called a risk allele the greater the OR is, the higher risk the allele will confer to the phenotype
what does an OR over under 1 mean?
potentially decreases the risk –> called protective allele
the smaller OR is, the lower risk the allele will confer to the phenotype
how is the OR calculated?
(number of T allele with persistent infection divided by number of T allele with cleared virus) DIVIDED BY (number of C allele with persistent infection divided by number of C allele with cleared virus)
what is a 95% Confidence interval?
statical probability for OR
the standard error of OR
what does it mean if 95% CI > 1?
significant risk effect
what does it mean if 95% CI contains 1?
no statistical signficance
what does it mean if 95%CI is under 1?
significant protective effect
why is a false positive likely?
due to a large number of tests
many SNPS are associated with a phenotype just by chance
the more SNPs tested, the higher the probability for false positive
what is the bonferroni correction?
corrected P=0.05/N (total number of SNPs tested)
what number is used for GWAS P value?
5x10^-8
what do human clinical trials compare to?
standard of care aka positive control
cannot use a negative control due to ethical reasons
what is the dynamic range?
the upper and lower limit of the data set
to have reliable results, what is essential?
having a large sample size
