Lecture: Diversion: Repetitive elements

Loading...

Из курса от партнера Johns Hopkins University

Algorithms for DNA Sequencing

367 оценки

Johns Hopkins University

Algorithms for DNA Sequencing

367 оценки

Курс 4 из 8 — Specialization Genomic Data Science

We will learn computational methods -- algorithms and data structures -- for analyzing DNA sequencing data. We will learn a little about DNA, genomics, and how DNA sequencing is used. We will use Python to implement key algorithms and data structures and to analyze real genomes and DNA sequencing datasets.

Из урока

Preprocessing, indexing and approximate matching

In this module, we learn useful and flexible new algorithms for solving the exact and approximate matching problems. We'll start by learning Boyer-Moore, a fast and very widely used algorithm for exact matching

Week 2 Introduction 1:34

Lecture: Boyer-Moore basics8:50

Lecture: Boyer-Moore: putting it all together6:15

Lecture: Diversion: Repetitive elements5:02

Practical: Implementing Boyer-Moore 10:01

Lecture: Preprocessing7:02

Lecture: Indexing and the k-mer index10:43

Lecture: Ordered structures for indexing8:06

Lecture: Hash tables for indexing7:13

Practical: Implementing a k-mer index 7:23

Lecture: Variations on k-mer indexes9:25

Lecture: Genome indexes used in research9:34

Lecture: Approximate matching, Hamming and edit distance6:53

Lecture: Pigeonhole principle6:37

Practical: Implementing the pigeonhole principle 9:30

Познакомьтесь с преподавателями

Ben Langmead, PhD
Assistant Professor
Computer Science
Jacob Pritt

Department of Computer Science

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.

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.

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.

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.

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.

Ознакомьтесь с нашим каталогом

Присоединяйтесь бесплатно и получайте персонализированные рекомендации, обновления и предложения.

Coursera делает лучшее в мире образование доступным каждому, предлагая онлайн-курсы от ведущих университетов и организаций.

© Coursera Inc., 2019 Все права защищены.

Загрузить из App Store

Загрузить в Google Play