Welcome to Week 3 of Statistics for Genomic Data Science.
Last week, we talked a little bit about linear modeling, and preprocessing and
normalization.
This week we're going to be talking a little bit more about modeling.
Finishing up some of the more technical details.
And we're also going to start talking about statistical significance.
And in particular, if you know anything about statistics, or
you've ever taken a statistics class before, you've heard about the P-value.
It's the most commonly used statistic in the world.
Millions and millions of P-values are calculated every year.
We're going to talk about how do you use those.
How do you perform hypothesis tests to try to make discoveries in genomics?
When you're doing statistical significance in genomics it gets a little bit tricky
though, because there's often many, many, many things that you're testing.
And you're trying to discover sort of a few signals among
many many different things that you might be trying to study.
And so we're going to talk a lot about multiple comparisons, and
how do you check to make sure that the discovery you're making is
likely to be real and not a false positive that came up just because you're doing so
much data sifting when you're looking through the data.
So we're going to talk about statistical significance and
multiple testing this week.
Jeff Leek, PhDAssociate Professor, Biostatistics
