0:03
This lecture is a brief diversion where we'll discuss repetitive elements,
what they are, and why they occur.
And why they're something that we are going to want to keep in mind as we
discuss the read alignment and assembly problems.
So a human genome sequence that we observed today is the end result of
a pretty complex evolutionary process.
And that process tends to introduce certain kinds of patterns into the genome.
Patterns that make the genome sequence very different from the kind of sequence
that you would get if you generated it randomly.
Like for example, by flipping a coin over and over again, or
by using a Python pseudorandom generator function to generate a random genome.
Maybe the most striking example of this fact is that
the human genome is extremely repetitive.
It's far more repetitive than a random string would be.
And so what do what I mean by this?
Well some pretty interesting things happen to genomes over time.
One thing that happens is that the genome gets invaded and infiltrated
by little bits of DNA, sort of DNA interlopers called transposable elements.
And these are tiny little chunks of DNA that are capable of getting themselves
incorporated into the genome.
And then when conditions are right, they're also capable of copying and
pasting themselves, or cutting and pasting themselves, throughout the genome.
1:29
So this diagram shows one such example of how a cut-and-paste mechanism happens,
how a transposable element can cut and paste itself throughout the genome and
it does this with the help of an important protein called a transposase.
So this kind of thing has happened many,
many times throughout evolutionary history to the human genome to the point where
about 45% of the human genome sequence is covered by transposable elements.
45% of all the bases in human genome came from a transposable element, and
there are many different kinds of transposable elements,
as you can see from this pie chart here.
They come in various abundances.
And then, there's one particularly renowned example of
a transposable element which is called Alu,
which is highlighted right here in this red bit of the pie chart.
So Alus actually occur many times in the genome, more than a million times.
And about 11 or so
percent of the human genome sequence is covered by these things called Alus.
So why should we care about these repetitive portions of the genome?
Do they really affect whether we can design good algorithms for
solving the read alignment problem and the assembly problem?
Well, actually they do.
So they do because they create ambiguity.
And if there are, for example 1 million copies of the Alu element, and one of our
sequencing reads comes from inside one of those copies of the Alu element, and
then we want to solve our read alignment problem, you can imagine,
that's difficult.
There's ambiguity.
It's going to be very difficult to figure out exactly which copy of the Alu repeat
our sequencing read came from, or maybe even impossible.
3:20
So here's a illustration of this.
So here is actually a picture of all the chromosomes of the human genome, and
everywhere where you see glowing green on this picture,
that's a place where there are Alus in the genome.
So, they're spread all over.
And if we're in a situation, let's say,
now this is a schematic version of our genome that has some unique sequences,
some non-repetitive sequences which are shown in blue.
And it also has many copies of the same repeat.
Maybe it's the Alu repeat, but those copies are shown in red.
And let's say we get some sequencing reads from the blue portions of the genome and
we try to solve the read alignment problem.
Well, that's good because we'll be able to resolve where they come from because they
come from non-repetitive portions of the genome.
4:08
Now, let's say we have a read that comes from inside one of the copies of this
repetitive element in the genome.
And this occurs in several copies, including the three that you see here.
Well we're not really going to have any idea whether the read came from the copy
that it truly came from, or whether it came from one of the other copies.
There's just going to be inherent ambiguity in the read alignment problem in
that sense.
4:32
So in our puzzle analogy, you could say this is a little bit like a puzzle
where half the pieces are just featureless blue sky, or something like that.
So that if you're trying to figure out where one of those pieces comes from,
it's going to be very difficult to distinguish which part of the sky it
actually comes from.
And then later when we talk about assembly, we'll see some other important
ways in which repetitive sequences in the genome are going to create problems for
our algorithms.
Ben Langmead, PhDAssistant Professor
Jacob Pritt
