5/6/25, 3:05 PM Chapter 1
Chapter 1 Code
Introduction to Genomics
Welcome to BioS 420-Genomics. My goals in this class are to
1. Familiarize you with the wealth of sequencing data available in public databases, how to find what you need and and how to use it.
2. Provide an understanding how genomes are sequenced and assembled
3. Give hands on experience in analyzing large sequencing data sets that provide mechanistic insights into a wide variety of biological fields
4. Provide a basic training and practical experience in command line and R programming
5. Develop confidence in using cutting edge open source bioinformatic software
What is a genome?
A genome was originally defined as the complete set of genes for an organism. (the -ome suffix indicates a complete set of the preceding parte of
the word. you will come across many (probably too many) “-omes” when studying modern biology)
But what is a gene?
During the history of the gene concept, genes have been conceptualized in various ways:
Genes as units of heredity for traits or phenotyes (usually just one)
Genes as “beads on a string”, linearly arranged sequences on a chromosome)
think recombination mapping, and recombination frequency
Genes as the instrcutions for “enzymes” (protein encoding)
Genes as functional strings of nucleotides
We will mostly talk about genes using the latter definition, and for most cases, all the other definitions also fit This being biology there are some
exceptions and complications … see [Link] ([Link] for an example of
the struggle
With the advent of widespread sequencing technology, we kind of ignore genes and define the genome as the complete set of nucleotides present
in an organisms chromosome(s).
Kinds of Genomes
Below is an example of a human genome, in the form of a chromosome preparation. Note the 22 pairs of linear genome, with variability at the last
pair.
Karyogram of a human genome From Genes and Disease. NCBI [Link]
([Link]
The arrangement of sequences in chromosomes can be surprisingly complex. Consider the platypus:
[Link] [Link] 1/11
5/6/25, 3:05 PM Chapter 1
Karyogram of a male platypus
From: [Link] ([Link]
While eukaryotes typically have a set of linear chromosomes, prokaryotes (bacteria and archaea) Typically organize therir genomes into circular
chromosomes. Very often bacteria have additional circular elements called plasmids that allow rapid transfter of functional DNA across species.
([Link]
From:
[Link]
([Link]
Eukaryotic cells also contain genomic elements out side of the main, nuclear chromosomes. These are the mitochondria and chloroplasts (in
plants). Interestingly, these genomes are circular, due to their origin from endosymbiosis of bacteria.
Viral genomes can take on several forms, being composed of RNA or DNA, which can be either single, or double stranded:
[Link] [Link] 2/11
5/6/25, 3:05 PM Chapter 1
Baltimore classification of viruses, based on type of molecule and strand information
From: [Link]
([Link]
Sizes of Genomes
Genomes come in a variety of sizes. For example the length of the haploid genome of humans is about 3 billion base pairs, or you will see it
referred to as 3 Gigabases (Gb).
The human genome is by far not the largest genome on record. As of May 2024, the largest sequenced genome on record is from a Fern, weighing
in at 160 Gigabases (see [Link] ([Link] )
The largest animal genome so far known is from a lungfish, at about 91 Gigabases ([Link]
([Link] This is 3 times larger than the 30 Gb genome of the world’s perfect animal, the Axolotl, which
should tell you that genome size isn’t everything.
An axolotl
Credit Th1098 ([Link]
[Link] ([Link]
Gene content variation across genomes
Similarly, the number of genes in a genome can be quite variable, and does not necessarily correlate strongly with either size of the genome, or
perceived complexity of the organism.
[Link] [Link] 3/11
5/6/25, 3:05 PM Chapter 1
log-log plot of genome size to number of genes. Note that there are many examples that do not fit a linear relationship
Credit: Smsaladi ([Link]
[Link]
([Link]
Table 1: Comparative genome sizes of humans and other model organisms[2] ([Link]
Organism Estimated size (base pairs) Chromosome number Esti
Human (Homo sapiens) 3.1 billion 46 25,0
([Link]
Mouse (Mus musculus) 2.9 billion 40 25,0
([Link]
Bovine (Bos taurus) ([Link] 2.86 billion[6] 60[7] 22,0
([Link] ([Link] (https
6) 7) 8)
Fruit fly (Drosophila melanogater) 165 million 8 13,0
([Link]
Plant (Arabidopsis thaliana 157 million 10 25,0
([Link]
Roundworm (Caenorhabditis elegans 97 million 12 19,0
([Link]
[Link] [Link] 4/11
5/6/25, 3:05 PM Chapter 1
Yeast (Saccharomyces cerevisiae 12 million 32 6,00
([Link]
Bacteria (Escherichia coli 4.6 million 1 3,20
([Link]
Credit: Wikipedia [Link] ([Link]
What accounts for genome size variation?
We’ve seen that genome size does not scale precisely with number of genes. A major reason for this is variation in the part of the genome that
does not code for proteins or functional RNAs. A major component of this are transposable elements (or transposons). Aproximately 45% of the
human genome is made up of copies of different transposons. Retrotransposons are particularly important as they are very prone to amplifying and
increasing over time if not checked:
Schematic of a retrotransposon copying itself to another place in the genome
Credit: Mariuswalter ([Link] [Link]
([Link]
This results in a lot of very similar and identical sequence throughout the genome, which as we will see can cause problems for sequencing and
assembly of genomes.
Other repetitive elements, such as direct repeats, where a short sequence is repeated hundred or thousands of times also play significant roles.
Direct repeats like ATAGATAGATAGATAG…etc…. make up about 3% of the human genome. These have a tendency to expand, as DNA
polymerase can sometimes lose its place and add extra copies of the repeat. This is a problem also in sequencing, as DNA polymerases are an
important part of most sequencing technologies.
In general the non-coding parts of the genome can grow and shrink due to errors in replication and transposon invasion more easily, as they are
less constrained than regions that code for protein, where keeping reading frames intact, and the downstream function of the peptide are
important, and selection removes more of the variability.
A quick refresher on gene structure
Prokaryotes and eukaryotes have some fundamental differences in how protein coding genes are arranged in the genome. In prokaryotes protein
coding sequences are uninterrupted open reading frames (ORFs), meaning that starting from the start codon (usually ATG, which will code for
Methionine in the RNA) there are no stop codons until the end of the coding unit. Typically several ORFs are arranged in an operon, which will
produce a single mRNA that encodes multiple proteins. Very often the genes in an operon have related functions. The regulatory sequences are
also directly flanking the expressed region. Hopefully I am bring up fond memories of the lac operon from your genetics class… Below is a more
generic model:
[Link] [Link] 5/11
5/6/25, 3:05 PM Chapter 1
credit: Thomas Shafee [Link]
([Link]
Eukaryotic genes can be much more complex. Most coding regions are interrupted by non coding stretches called introns. The coding portions are
called exons. What’s more, there can be multiple ways of including or excluding exons into the final messenger RNA, which leads to many
different proteins (called isoforms) being produced from a single gene. The regulatory sequences are more complex and can be at various
positions, and even very far away in the genome.
credit: Hoffmeier [Link]
([Link]
A quick refresher on Evolution
All living things share common ancestors with all other living things on earth, in the more or less distant past. Typically, the more ancient the
common ancestor, the less similar the two species appear, and vice versa. Intuitively, we can see that we as humans share many unique features
with apes (biggish brains, extremely reduced tails, opposable thumbs etc…). Similarly rodents like capybara and mole rats share a suite of features
(especially those teeth) with each other, that are not found in apes. But we mammals still share a significant number of traits (milk, hair, live birth)
that are not found in other vertebrates, and vertebrates share critical unique features not found in other animals (vertebral column, skulls, internal
skeletons, tails (extremely reduced secondarily in apes)). Consistent with this, all evidence points to humans and chimps sharing a very recent
common ancestor, all of us apes sharing a common ancestor with rodents more millions of years ago, and all mammals sharing a common
ancestor with fish even longer ago.
By looking at groups of related organisms, we can infer some traits of their common ancestor (which of course no longer exists). For example the
common ancestor of apes and rodents in all likelihood was warm blooded, fed its young with milk, was covered in hair, and gave live birth. We infer
this because it is very unlikely that this suite of complex traits would have evolved together more than once. With the information we have here we
[Link] [Link] 6/11
5/6/25, 3:05 PM Chapter 1
cannot say whether the common ancestor of mammals had a big brain, or had constantly growing teeth. This is because it is just as likely that
these traits evolved after the lineages diverged, or that they were present in the ancestor, and were lost in one of the lineages.
This history is reflected in our genomes. You may have heard that we share 99% (or 98, or 95) of our genome with chimapanzees. You may have
then thought to yourself, what does that mean? In the end probably not much, such comparisons are complicated. Below is a more specific
comparison:
From: [Link] ([Link]
Below is an example of how the divergence between chimps and humans can look at the level of bases in a coding gene. Can you spot the
differences?
[Link] [Link] 7/11
5/6/25, 3:05 PM Chapter 1
Overall the mRNAs are each 5786 bases long, and the sequences are identical at 5761 of them . Only a handful of the differences cause a change
in the amino acid sequence encoded. Despite this very high similarity, scientists have proposed that the human specific changes in this gene
(called FoxP2) were important for the evolution of language (see [Link] ([Link] One
line of evidence is that human mutations in this gene lead to difficulties in developing speech and language (apraxia). Another is that mice with
similar mutations also develop problems in vocalizations typical of rodents, showing that this gene has a long evolutionary history of being involved
in communication. Finally the authors of the linked study found there was evidence of strong selection around the human FoxP2 gene that was not
present in other apes.
Sounds good, but it is definitely not so simple, and the idea is still controversial. But it illustrates a couple of important concepts. One is that it
shows how genomic sequencing is revolutionizing all aspects of biology. For example, one piece of evidence that the human specific changes in
FoxP2 are not important was the finding that the same mutations exist in Neanderthal (i.e., they have sequenced the genome of an extinct species
well enough to know this: [Link] ([Link] , who were thought to be non-
verbal.
The other important idea is that of homology.
[Link] [Link] 8/11
5/6/25, 3:05 PM Chapter 1
This can be defined as: structures or traits appearing in a set of organisms that are derived from a similar structure or trait that was
present in the common ancestor of all of the organisms being compared.
Since common ancestors for species no longer exist, and features can evolve and differentiate quite quickly, homology can only be inferred by
comparing existing creatures (and fossils if available). Typically scientist use similarity in position, composition, and arrangement of sub-features as
evidence of homology. A classic example of this is the vertebrate forelimbs:
credit: Волков Владислав Петрович [Link] ([Link]
The arm of humans, front leg of dogs, wings of birds, and front flippers of whales, are all considered to homologous as forelimbs, and the common
ancestor of vertebrates most likely had a forelimb with all of the same elements (upper humerus, paired radius and ulna, carpals then phalanges)
that are found in all of these species, despite the quite drastic alterations that have occurred as the lineages have split and evolved. The alternative
hypothesis is that all of these bones with identifiable properties evolved very similar patterns of shape, arrangement, and placement on the body
independently multiple times, which seems quite unlikely.
The concept of homology can be applied to molecules as well. DNA and protein are very well suited for comparing across species. Below is an
alignment of the human FoxP2 gene sequence and its mouse homolog.
[Link] [Link] 9/11
5/6/25, 3:05 PM Chapter 1
As you can see, despite the passage of about 75 million years since mice and humans shared a common ancestor, the sequences line up pretty
well. Given the 4 base nature of the genetic code, there is a 1/4 chance that any nucleotide from one sequence will match with another, and a
string of matches would have a probability of 1/4^n, where n=equals the number of bases in the match. So, clearly these sequences are not similar
due to random chance. The best explanation for the similarity of these sequences is that the common ancestor of mice and humans had a FoxP2-
like gene, and that the specific sequence of most of the protein is very important.
For many genes it is relatively simple to identify potential homologs in different species: Take your gene of interest, search the genome of interest
and identify the most similar sequence there. If you then take this best match, and search for the best match, you should find the the same gene
you started with. This is referred to as searching for the reciprocal best match. This works pretty well for a lot of genes, and especially in closely
related species. But there are many reasons why you may not get an easy answer.
One is that not all genes are as constrained in evolution, and random mutation accumulation over time can erase the initial similarity pretty quickly.
Some genes encode proteins where only a small functional domain is important, and the rest of the sequence is less selected on, leading to rapid
diversification.
[Link] [Link] 10/11
5/6/25, 3:05 PM Chapter 1
Another common issue is caused by the tendency of genes to duplicate in evolution. Sometimes DNA polymerase slips, or there are defects during
meiosis, leading to chunks of DNA being duplicated, sometimes including a whole functional gene. Sometimes, the whole genome duplicates and
every gene has a new copy. These new copies can evolve differently in the descendants of the original species where the duplication occurred, in
ways that obscure the original relationships. This can lead to situations where the reciprocal best match test doesn’t work. A common example is
that one species had a duplication event, and the other didn’t, and one duplicated copy evolved more rapidly or in a different direction than the
other.
We will get into the details about how to identify likely homologous genes in later chapters. But I want to emphasize the importance of being able to
identify homologs, particularly in regard to human health (this is of course not the only important thing). If two sequences had a common ancestor,
the ancestral sequence likely had a suite of important functions. It is likely that most of these function will be maintained in some form in its
descendant sequences. Thus if we can understand the function of a molecule in one species, we have a strong basis for predicting the functions in
another species. This is the main basis and value of using model organisms to help us understand human biology.
[Link] [Link] 11/11