We noted you are experiencing viewing problems
-
Check with your IT department that JWPlatform, JWPlayer and Amazon AWS & CloudFront are not being blocked by your network. The relevant domains are *.jwplatform.com, *.jwpsrv.com, *.jwpcdn.com, jwpltx.com, jwpsrv.a.ssl.fastly.net, *.amazonaws.com and *.cloudfront.net. The relevant ports are 80 and 443.
-
Check the following talk links to see which ones work correctly:
Auto Mode
HTTP Progressive Download Send us your results from the above test links at access@hstalks.com and we will contact you with further advice on troubleshooting your viewing problems. -
No luck yet? More tips for troubleshooting viewing issues
-
Contact HST Support access@hstalks.com
-
Please review our troubleshooting guide for tips and advice on resolving your viewing problems.
-
For additional help, please don't hesitate to contact HST support access@hstalks.com
We hope you have enjoyed this limited-length demo
This is a limited length demo talk; you may
login or
review methods of
obtaining more access.
Printable Handouts
Navigable Slide Index
- Introduction
- Outline
- Statistical associations vs. causal effects
- What is a causal effect?
- Causal effects and counterfactuals
- Defining causal effects at the population level
- A way to estimate average causal effect (ACE)
- Causal effect in the exposed subpopulation
- The “naive” association analysis
- Counterfactual outcomes in different settings
- Classical regression estimates vs causal effects
- Adjustment for true confounders
- Adjustment may sometimes make things worse
- Direct vs. indirect effects
- Mendelian randomization-genes as variables
- General instrumental variables estimation (1)
- General instrumental variables estimation (2)
- General instrumental variables estimation (3)
- Mendelian randomization example
- IV estimation in R (using library(sem))
- IV estimation - testing untestable assumptions
- Association structure (1 genotype, 2 phenotypes)
- One genotype and two phenotypes
- Can we test pleiotropy
- Summary
- References
- Thank you
Topics Covered
- Definitions of causal effect
- Causal graphs, confounding and adjustment
- Instrumental variables estimation
- Mendelian randomization
- Association structure with one genotype and two phenotypes
Talk Citation
Fischer, K. (2017, March 29). Causal inference in genetic epidemiology: Mendelian randomization and beyond [Video file]. In The Biomedical & Life Sciences Collection, Henry Stewart Talks. Retrieved December 6, 2024, from https://doi.org/10.69645/XDOZ9569.Export Citation (RIS)
Publication History
Financial Disclosures
- Prof. Krista Fischer has not informed HSTalks of any commercial/financial relationship that it is appropriate to disclose.
Other Talks in the Series: Statistical Genetics
Transcript
Please wait while the transcript is being prepared...
0:00
My name is Krista Fischer.
And I'm working as a Senior Researcher
at Estonian Genome Center,
University of Tartu, Estonia.
My background is in biostatistics
and in genetic epidemiology.
My research interest
for about two decades
has been causal inference,
and in recent years,
it has been more
in the field of genetics,
so Mendelian randomization
is one of my favorite topics of research
and also to teach.
0:31
Today, we're going to talk
about causal inference,
causal effects, but causal effect
may be simple in everyday language,
but it's not so easy
to define it mathematically.
I introduced one possible way
to talk about causal effects,
so called language
of counterfactual variables.
Then I will briefly speak
about causal graphs
and show some examples
related to confounding
and adjustments because often
we need to answer questions
whether we should adjust
for some confounders,
potential confounders or not.
And then I'll move
to Mendelian randomization topics,
specifically,
instrumental variables estimation,
so it can be used in more general cases
in Mendelian randomization.
And then we'll see
what Mendelian randomization is.
so it can The only possible way
to analyze association structures
with one genotype and two phenotypes,
as you see there are actually
more possible association structures
and Mendelian randomization
can be used only
some assumptions are fulfilled.
Then we will discuss these assumptions
and ways to do checks,
and we'll end with some references.
Hide