<-chapter_06|Chapter 06^table_of_contents|Table of Contents^chapter_08|Chapter 08-> Chapter 07. Analysis of %%gene sequences%% Although eukaryotic genes may be generally more interesting to most students, it is useful to first consider bacterial genes. Most eukaryotic molecular biologists use bacteria as tools for various things (e.g., molecular cloning; see [[chapter_09|Chapter 09]]), so it’s useful to understand how bacteria work from a practical perspective. Also, although bacterial genes have some pretty important differences compared to eukaryotic genes, many basic principles are the same. ===== Bacterial gene structure ===== This is what a typical bacterial gene might look like schematically.

{{ :bacterial_gene_structure.png?400 |}} Figure 7.1. Anatomy of a bacterial gene. See Table {{ref>Tab1}} for descriptions of different parts of this gene. The straight horizontal line represents double-stranded DNA, whereas the wavy line represents mRNA. Note that the mRNA is slightly longer than the coding sequence.

^ DNA or RNA element ^ Function ^ | Promoter | To target RNA polymerase to DNA and to start transcription of a mRNA copy of the gene sequence. | | Transcription stop (terminator) | To instruct RNA polymerase to stop transcription. | | Shine-Dalgarno sequence and start codon | Shine-Dalgarno sequence in mRNA will load ribosomes to begin translation. Translation almost always begins at an AUG codon in the mRNA (an ATG in the DNA becomes an AUG in the mRNA copy). Synthesis of the protein thus begins with a methionine (see Fig. 6.4). | | Coding sequence | Once translation starts, the coding sequence is translated by the ribosome with the help of transfer RNAs (tRNAs), which read three bases at a time in consecutive sequence (one after the other with no overlap). Amino acids will be incorporated into the growing polypeptide chain according to the genetic code (see Fig. 6.4). | | Translation stop | When one of the three stop codons (UAG, UGA, or UAA; see genetic code in Fig. 6.4) is encountered during translation, the polypeptide will be released from the ribosome. |

Table 7.1. Bacterial gene parts list.

For example, a bacterial gene coding sequence that is 1,200 nucleotide base pairs in length (including the ATG start codon but not including the stop codon) will specify the sequence of a protein 1200/3 = 400 amino acids long. Since the average molecular weight of an amino acid is 110 Daltons (1 Da = 1 g/mol), this gene encodes a protein of about 44 kDa - the size of a pretty average protein. There are also important cis-acting regulatory elements that are outside of the coding sequence of a gene but nonetheless are important for a gene's function (discussed further in [[chapter_10|Chapter 10]]). ===== Identifying a gene based on DNA sequence data ===== In earlier chapters we discussed how genes are classically identified by their function. That is, the existence of the gene is recognized because of mutations in the gene that give an observable phenotypic change when compared to wildtype. Historically, many genes have been discovered because of mutations and their effects on phenotype. Genes were then mapped and/or cloned (and therefore physically identified) by complementation ([[chapter_09|Chapter 09]]). Now in the era of large-scale and inexpensive genomic sequencing, many genes of no known function can be detected by looking for patterns in DNA sequences. The simplest method that works for bacterial genes (but not for most eukaryotic genes, as we will see later) is to look for stretches of DNA sequence that lack stop codons. These sequences are known as open reading frames or ORFs. This method works because a truly random sequence should contain an average of one stop codon in about every 21 codons (i.e., there are a total of 64 possible codons and there are 3 stop codons). Thus, the probability of a random occurrence of even a short open reading frame of (for instance) 100 codons (or 300 base pairs (bp) of DNA) without a stop codon is relatively small: $p = (\frac{61}{64})^{100} = 8.2 \times 10^{-3}$, or slightly better than a 1 in 10,000 chance. Since the genetic code is a triplet code, one would need to search through three possible reading frames for ORFs. Identifying genes in DNA sequences from eukaryotes is usually more difficult than in bacteria. First, this is because gene coding sequences in eukaryotes are separated by long sequences that do not code for proteins called non-coding intergenic regions. In other words, eukaryotic genomes are usually less gene-dense than prokaryotes, so there is a lot more DNA sequence to search through. Moreover, the coding sequences of most eukaryote genes are interrupted by introns, which are sequences that are removed (spliced) from the mRNA before translation. In other words, eukaryotic genes at the DNA level are not co-linear with proteins. The presence of introns breaks up the open reading frames into short segments called exons; exons are joined together during splicing to making genes much harder to distinguish from non-coding sequences. The process of splicing removes introns and joins exons to form mature mRNA. This makes computer programs that search for ORFs much less useful when trying to find eukaryotic genes from a large genomic DNA sequence. See [[chapter_12|Chapter 12]] for details on splicing. The maps below in Fig. {{ref>Fig2}} show 50 kbp (1 kbp = 10³ bp) segments of genomic DNA from yeast, Drosophila, and humans. The dark gray boxes represent exons, and the light gray boxes represent introns. The boxes above the line are transcribed from left to right and the boxes below are transcribed from right to left. Remember that DNA is double-stranded and coding sequences are read in a 5’ to 3’ direction, so genes can exist on either strand of DNA but in opposite directions. Names have been assigned to each of the identified genes. Although the yeast genes contain few introns and are packed closely together, the Drosophila and human genes are spread apart and interrupted by many introns. Sophisticated computer algorithms and/or experimental approaches are used to identify these dispersed gene sequences. We will examine eukaryotic gene structure more closely in [[chapter_13|Chapter 13]].

{{ :eukaryotic_gene_structure_comparison.png?400 |}} Segments of physical gene maps from yeast, Drosophila, and humans. Using yeast as an example: labels that follow standard yeast nomenclature conventions (e.g., //RGD2//, //FET5//) have likely been studied and are curated in the yeast genome database. By comparison, the code-like letters and numbers (such as YFL046W) are just code names for predicted or poorly characterized genes for which their function is probably not yet known. See text for other details.

An experimental way to identify eukaryotic genes physically is by examining mRNA instead of DNA. If an mRNA exists in a cell, this means that it was most likely transcribed from a gene. mRNAs can be purified from cells biochemically, and an enzyme called reverse transcriptase (usually isolated from various types of retroviruses) can be used to convert mRNAs into DNAs called complementary DNAs (cDNAs). cDNA does not exist in nature - it is created by scientists in the lab. We can sequence (discussed below) these cDNAs and compare the sequences to genomic DNA sequences to identify and locate genes. Randomly sequenced cDNAs derived from mRNAs isolated from cells or tissues are often called expressed sequence tags (ESTs). If an EST sequence matches small stretches of genomic DNA sequence, that would suggest that a gene is there even if we don't know what the gene or gene product (protein) does. There are now more modern technologies, such as RNAseq (see below), that are much faster than traditional cDNA-based approaches - the details are not important for now, but RNAseq is very similar to the NGS technologies introduced below. This book is more focused on how to conceptually analyze gene function rather than details of molecular and genomic approaches to identifying genes. ===== How to sequence DNA: background information ===== Reading the order of nucleotides from a piece of DNA is called DNA sequencing. To see how gene sequences from DNA are actually obtained, we will first need to revisit some fundamentals of the chemical structure of DNA and how DNA replication works (see also [[chapter_06|Chapter 06]]). DNA is double-stranded (this is commonly abbreviated as dsDNA). Each strand of DNA is directional - the two ends of a single strand of DNA (ssDNA) are structurally different. The different ends are usually called the 5’ and 3’ ends. This refers to different positions on the ribose sugar ring where the linking phosphate residues attach (Fig. {{ref>Fig3}}).

{{ :dna_chemical_structure.jpg?400 |}} Chemical structure of DNA revisited. The "phosphate-deoxyribose backbone" in the diagram is also called the phosphodiester backbone. It can also be helpful to revisit Fig. 6.1. Source: [[https://commons.wikimedia.org/wiki/File:DNA_chemical_structure.svg|Wikimedia]]. Credit: M.P. Ball. Licensing: [[https://creativecommons.org/publicdomain/zero/1.0/deed.en|CC0 1.0]].

In a double-stranded DNA molecule, the two strands run anti-parallel to one another and are held together by base pairs through hydrogen bonds. Fig. {{ref>Fig3}} is structurally correct but difficult to conceptualize when thinking about DNA replication. A more abstract model of DNA structure that is more useful for thinking about DNA replication is shown in Fig. {{ref>Fig4}}:

{{ :dna_structure_simple.jpg?400 |}} conceptual diagram showing the antiparallel and base pairing nature of dsDNA. Credit: M. Chao.

DNA polymerase is the key enzyme in the DNA sequencing methods we will be considering. The general reaction carried out by DNA polymerase is to synthesize a copy of a DNA template using the chemical precursors dATP, dGTP, dCTP, and dTTP (deoxyribonucleotide triphosphates, collectively abbreviated as dNTPs). There are lots of different kinds of DNA polymerases, many with specialized functions or role. But all DNA polymerases have two fundamental properties in common: * New DNA is synthesized only by elongation of an existing strand at its 3’ end. * Synthesis requires three things: dNTPs, a template strand, and an existing nucleic acid with a free 3’ hydroxyl (-OH) end base paired to the template strand called a primer. A general substrate for DNA polymerase looks like Fig. {{ref>Fig5}}:

{{ :dna_replication_substrate.jpg?400 |}} The substrate for DNA polymerase includes a template strand, a primer with a free 3' hydroxyl group, and dNTPs (not shown). The primer can be made from single-stranded DNA (ssDNA) or RNA, but in this diagram the primer is ssDNA. Credit: M. Chao.

Note that the template strand can be as short as 1 base or as long as several thousand bases. The duplex region (i.e., the double stranded region) of this substrate has to be just large enough for DNA polymerase to bind to it. As a general rule of thumb, this region should be at least around 20 bp in length (it can be much longer but not shorter). After addition of DNA polymerase and dNTPs, a single strand of DNA will be synthesized, starting from the 3' hydroxyl group of the primer and progressing in a 5' to 3' direction. This direction of synthesis is relative to the new strand being made, not the template strand (Fig. {{ref>Fig6}}).

{{ :dna_synthesis.jpg?400 |}} DNA synthesis by DNA polymerase occurs in a 5' to 3' direction and initiates off of the 3' hydroxyl group of the primer. Credit: M. Chao.

===== DNA Sequencing: the details ===== We first discuss an older but still relevant type of DNA sequencing technology called Sanger sequencing. The basic method was invented in 1977 and is named after its inventor [[wp>Frederick_Sanger|Frederick Sanger]]. Consider a segment of dsDNA that is about 1000 base pairs long that we wish to sequence. To sequence this DNA, we first need to have a source of DNA material, which we consider [[chapter_07#Polyermase_Chain_Reaction_(PCR)|further below]]. Assuming we have several μg (microgram; 1 μg = 10^-6 g) of DNA template to sequence, we first separate the two DNA strands by heating the DNA to about 100 °C to melt the hydrogen bonds that hold the ssDNAs together through base pairing. Next, a short single-stranded primer (about 18-20 bases long) designed to be complimentary to the end of one of the strands is allowed to anneal to the single stranded DNA. These primers are designed with the help of a computer and synthesized through commercially available services. The primer is added at a huge molar excess compared to the DNA you are trying to sequence – so most ssDNAs will pair with primer DNA rather than their original complementary partner ssDNA. The resulting DNA hybrid looks much like the general DNA polymerase substrate shown in Fig. {{ref>Fig5}}. DNA polymerase is then added along with the four dNTP nucleotide precursors (dATP, dGTP, dCTP, and dTTP). A small quantity of a slightly different nucleotide precursor called a dideoxyribonucleotide triphosphate is also added. Dideoxy nucleotide precursors are abbreviated ddATP, ddGTP, ddCTP, and ddTTP (or ddNTPs collectively). The ddNTPs have each been chemically labeled with a unique fluorophore that emits a different color of light after stimulation with a laser – for instance, green for ddATP, cyan for ddCTP, yellow for ddGTP, and red for ddTTP. These molecules are identical to the normal dNTPs in all respects except that they lack a hydroxyl group at their 3’ position (3’ OH) and that their nucleotide bases are chemically labeled with a fluorophore (Fig. {{ref>Fig7}}; fluorophore not shown). ddNTPs are also called chain terminators.

{{ :ddntp.jpg?400 |}} Structural comparison between ddNTPs and dNTPs. ddNTPs do not have a hydroxyl group on the 3' carbon of the ribose ring. Therefore, once a ddNTP is added to the end of a newly synthesized DNA strand, that strand cannot be polymerized further by DNA polymerase. Source: [[https://openstax.org/books/microbiology/pages/12-2-visualizing-and-characterizing-dna-rna-and-protein|OpenStax Microbiology]]. Licensing: [[https://creativecommons.org/licenses/by/4.0/deed.en|CC BY 4.0]].

The polymerase reactions are initiated, usually by moving the reaction mix from an ice bath to a 37 °C bath. After the DNA polymerase reactions are complete, the samples are denatured with heat again and separated using an electrical field through a high-resolution gel system that allows DNA strands of different lengths to be resolved; even differences of 1 nucleotide can be resolved((This kind of gel system is made of a material called polyacrylamide. The technique of "polyacrylamide gel electrophoresis" is commonly abbreviated as PAGE.))! The DNA sequence can be read from the gel by scanning it with a laser and using a light detector to visualize different colors of fluorescent light (Fig. {{ref>Fig8}}).

{{ :sanger_sequencing.png?400 |}} Automated Sanger sequencing. On the left side of the image, each duplex containing a terminated chain is a different DNA molecule; this is why a large quantity of starting DNA is needed for Sanger sequencing. The chains are terminated randomly, but the diagram shows the terminated chains in order of size for illustrative purposes. Note the order of the colored bands – compare the lengths of the randomly terminated strains on the left with the order of migration on the gel on the right. Separating them on a gel based on size orders the bands such that the ddNTPs reveal the DNA sequence. Credit: M. Chao.

ddNTPs can be incorporated into DNA, but once a ddNTP has been incorporated, further elongation stops because the resulting DNA will no longer have a free 3’ OH end. Each of the ddNTPs is added at about 1% the concentration of the normal nucleotide precursors. Thus, using ddATP as an example, about 1% of the elongated chains will randomly terminate at the position of an A in the sequence; the same will be true for the other ddNTPs. Once all the elongating chains have been terminated, there will be a population of newly synthesized and fluorescently labeled ssDNA strands that have terminated at the position of the sequence. The right-hand side of Fig. {{ref>Fig8}} (the part with colorful horizontal bands) represents a gel where DNA fragments of different lengths, each ending with a chain terminator, are separated using electrophoresis on a high-resolution gel. The DNA fragments migrate toward the positively charged cathode (because the phosphate groups on the backbone of DNA are negatively charged), and shorter fragments migrate faster than longer fragments; this technique is called electrophoresis. A laser and light detector coupled to a computer then automatically reads the different colored bands in order and determines the DNA sequence (Fig. {{ref>Fig8}}). ===== Polymerase Chain Reaction (PCR) ===== Now let’s consider how to physically obtain DNA for sequencing. A relatively large amount of DNA (approx. 1 μg worth for a piece of DNA several kbp long) is needed for the Sanger chemistry to work. As a student, you might not have a feel for how much DNA this is, but it's a substantial amount! To give you an idea of the scale of the problem, see [[chapter_07#Questions_and_exercises|Exercise 1 below]]. In the early days of molecular biology research, DNA for sequencing was obtained from cloned DNA segments, which can be difficult to create, but once created can easily be produced to give the quantity of DNA needed (molecular cloning is still used today for various purposes). We will discuss some methods for cloning new genes in [[chapter_09|Chapter 09]]. If we want to quickly find the sequence of a new mutant allele of a known gene, we need an easy way to obtain a relatively large quantity of this DNA without needing to go through molecular cloning. The easiest and most common way to do this is to use an //in vitro// method known as the polymerase chain reaction (PCR) that was developed by [[wp>Kary_Mullis|Kary Mullis]] in the mid-1980s (Fig. {{ref>Fig7}}). The steps in PCR are as follows: - A crude preparation of chromosomal DNA is extracted from the tissue source of interest (there is usually not enough DNA for sequencing from this step). - Two short primers (each about 18-20 bases long) are added to the DNA at an enormous molar excess. The primers are designed from the known genomic sequence to be complimentary to opposite strands of DNA and to flank the chromosomal segment of interest. - The double stranded DNA is melted by heating to around 100 ˚C (in practice we usually use 95 °C) and then the mixture is cooled (usually around 50 °C) to allow the primers to anneal to the template DNA. Since there is a huge molar excess of primer vs. template, most of the template will anneal with primer rather than re-anneal with its original partner strand. - DNA polymerase and the four nucleotide precursors are added, and the reaction is incubated at around 72 ˚C for a period of time to allow a copy of the segment to be synthesized. The reason we use 72 °C instead of 37 °C like we do for most enzymatic reactions is that we use a special heat stable enzyme called Taq DNA polymerase instead of standard DNA polymerase. - Repeat steps 3 and 4 multiple times (up to 30-35 cycles). To avoid the inconvenience of having to add new DNA polymerase in each cycle (due to the heating cycle eliminating DNA polymerase activity), a special DNA polymerase called Taq polymerase that can withstand heating to 100 ˚C is used. The idea behind PCR is that in each cycle of melting, annealing, and DNA synthesis, the amount of the DNA segment is doubled. This gives an exponential increase((If you think about it, the idea behind PCR is actually quite simple. It is mimicking how DNA replication occurs in dividing cells. Dividing cells with unlimited resources also replicate exponentially.)) in the amount of the specific DNA bounded by the primers on either side as the cycles proceed. After 10 cycles the DNA is amplified 2¹⁰-fold (2¹⁰ = 1024; in other words, about 1000-fold) and after 20 cycles the DNA will be amplified 2²⁰-fold (or approximately 10⁶-fold). Amplification usually continues until all of the nucleotide precursors are incorporated into synthesized DNA. The resulting PCR product can be used for DNA sequencing.

{{ :pcr.jpg?400 |}} Polymerase chain reaction (PCR). See text for details. Source: [[https://www.genome.gov/genetics-glossary/Polymerase-Chain-Reaction|National Human Genome Research Institute]]. Licensing: public domain.

Let's do some quick back of the envelope math to see if the numbers add up. Let's assume we start with a single molecule of dsDNA 1000 bp long that we want to sequence. Our goal is to get 1 μg of this DNA to use in Sanger sequencing. How many cycles of PCR do we need to do to get this amount of DNA? The first thing we need to do is to figure out how many molecules is 1 μg of a 1000 bp fragment of dsDNA. From a quick Google search, we learn that the average molecular weight of a nucleotide is approximately 330 Da (g/mol). Since DNA is double-standed, this means the the average molecular weight of a base pair is 660 g/mol, and the approximate molecular weight of a 1000 bp dsDNA fragment is therefore: $$660 \times 1000=6.6 \times 10^5 \text{ g/mol}$$ 1 μg of this DNA is equivalent to: $$\frac{1 \times 10^{-6} \text{ g}}{6.6 \times 10^5 \text{ g/mol}} = 1.5 \times 10^{-12} \text{ mol}$$ Using Avogadro's number we can convert this to number of molecules: $$1.5 \times 10^{-12} \text{ mol} \times 6 \times 10^{-23} \text{ molecules/mol} = 9 \times 10^{11} \text{ molecules}$$ Finally, we solve the exponential equation to find $n$, which is number of cycles: $$ 2^n = 9 \times 10^{11}\\ n\log{2} = \log{9} + 11\\ n = \frac{\log{9}+11}{\log{2}} \approx 38 $$ Starting with a single molecule of dsDNA, you would only need 38 cycles of PCR to obtain 1 μg of a 1 kb long DNA. In reality, you rarely start with just a single molecule of dsDNA, so 20-30 cycles of PCR is usually enough even when you have very little starting material. Each cycle of PCR takes approx. 1-2 minutes, so the entire procedure takes just 1-2 hours. The exponential amplification of PCR makes it relatively easy and straightforward to generate large quantities of a piece of DNA for various types of analysis, even if you have a small amount of starting material. For example, nearly all of the forensic DNA analysis that is dramatized on police procedural shows like NCIS or Criminal Minds is done using PCR in real life crime labs. PCR can also be used for genotyping human diseases or for paternity testing. ===== NGS technologies ===== ==== NGS vs Sanger sequencing ==== The Sanger sequencing method [[chapter_07#DNA_Sequencing:the_details|described above]] was used to complete the sequencing of the genomes of many organisms we have talked about (or will talk about) in this class, including the bacterium //Eschericha coli//, yeast, Drosophila, //C. elegans//, and even humans. This technology is still used today for certain purposes. However, next generation sequencing (NGS) technologies developed over the last two decades or so has made DNA sequencing significantly faster and cheaper. For instance, the original Human %%Genome%% Project, which started in 1990, took 13 years and thousands of researchers, and cost \$2.7 billion dollars to complete. With current NGS technologies, that cost has dropped to \$600 per genome as of this writing in 2023 and uses a fully automated machine that one technician can easily operate. A complete human genome can be sequenced now in just a few days. ==== Illumina sequencing ==== There are several different types of NGS technology. The most common type of NGS is called Illumina sequencing (Figs {{ref>Fig10}}-{{ref>Fig11}}), and the sequencing chemistry in this method is called sequencing by synthesis. Despite the fancy name it is not conceptually different than Sanger sequencing in that it still depends on DNA polymerase. The key difference is in how many templates are sequenced at once. Instead of sequencing one fragment of DNA at a time, Illumina sequencing takes an entire genome (for instance, from a cancer patient tumor biopsy sample) and fragments it into millions of small pieces, each around 300-600 bp long. These fragments are affixed to a flow cell - a device that is roughly the size of a microscope slide – and amplified in situ by a process called solid-phase bridge PCR. Each fragment, once affixed and amplified, has a unique position on the slide. In essence, you are forming a "colony" of DNA clones at different positions on the flow cell. This process is called cluster generation (Fig. {{ref>Fig10}}). The flow cell is then exposed to sequencing reagents similar to Sanger sequencing, except that instead of ddNTP chain terminators, fluorescently labeled dNTPs with a removable blocking group are used that pauses the elongation reaction each time a new nucleotide is added (Fig. {{ref>Fig11}}). Just like with Sanger sequencing, each added nucleotide is coupled to a fluorophore that emits a different color. A camera takes a picture of the entire slide, and a computer keeps track of the fluorescent signals at each unique position within the flow cell. The blocking groups and fluorophores are then chemically removed, the flow cell washed, and new sequencing reagents are added to "sequence" the next nucleotide. Each individual template is sequenced very slowly compared to Sanger sequencing. For instance, the speed of DNA polymerase is roughly 100 bp/sec - so that is how fast Sanger sequencing can go((Actually, the slowest step in Sanger sequencing is the separating of synthesized DNA fragments through a gel. The actual chemistry takes just 10 minutes, but preparing and running the gel can take several hours.)). In Illumina sequencing, adding one nucleotide, taking a picture, removing the fluorophores and blocking agents, and re-starting the sequencing reaction is slow - this step can take minutes. The trick is that Illumina sequencing is analyzing millions of fragments of DNA at the same time (it is analyzing multiple DNA fragments in parallel). Thus, on a strand-by-strand basis it is slow, but in terms of number of bases sequenced per unit time, it is orders of magnitude faster than Sanger sequencing.

{{ :cluster_generation.jpg?400 |}} Cluster generation for Illumina sequencing. The diagram shows some of the details of cluster generation. Steps (1)-(3) are known as bridging PCR. As a student, the important concept is that each cluster contains identical copies (or clones) of ssDNA that are affixed to a specific physical location on the surface of the flow cell. Credit: M. Chao.

A computer determines the DNA sequence of each fragment by analyzing the fluorescent signals during each cycle of sequencing and assembles the sequences from individual fragments into a complete sequence. The error rate in this process is relatively high compared to Sanger sequencing – on average, one error is called per every 1000 bases sequenced, but sequencing depth (that is, the number of times a DNA sequence exists in different fragments that are read) can overcome the error rate by reading the same sequence multiple times on different DNA fragments.

{{ :illumina_sbs.jpg?400 |}} Sequencing by synthesis in Illumina sequencing. The diagram shows a simplified schematic. The left side of the diagram is a side view of the flow cell, whereas the right side shows a top down view. Each DNA cluster only shows a single ssDNA template for simplicity, but in reality there can be many hundred copies in each cluster (see Fig. {{ref>Fig10}}). Also, each ssDNA template is usually 300-600 nt long but here a short ssDNA is shown for illustrative purposes. Sequencing originates off of a universal primer based on the red sequence in Fig. 7.8; in this diagram the sequencing primer is simplified as a 3 nucleotide sequence 5'-CGG-3'; in reality it will be much longer (~20-30 nt). Each type of nucleotide (G, A, T, or C) contains a fluorophore that emits a different color; it also contains a blocking group (black octagon) on the 3' carbon of the nucleotide. When the blocking group is removed, it regenerates a hydroxyl group so that the next round of sequencing can proceed. The diagram shows just two clusters for simplicity; in reality, each flow cell can contain millions of different clusters. A fluorescence microscope scans the flow cell after each sequencing reaction cycle and a computer keeps track of the results. Credit: M. Chao.

==== Other NGS technologies and applications ==== Technologies such as Illumina sequencing are now the preferred method for most types of large-scale DNA sequencing, and it has been adapted for related technologies such as RNA sequencing (RNAseq). In RNA sequencing, RNA is first converted to complementary DNA (cDNA) using an enzyme called reverse transcriptase; the cDNA is then sequenced using standard Illumina sequencing. The number of reads of a particular RNA sequence gives information as to how much a particular gene is expressed via transcription. Unlike DNA (which is present at 2 copies per cell), mRNAs can be present in hundreds of thousands of copies per cell; this means that Illumina sequencing is sensitive enough to sequence mRNA from single cells (single cell RNAseq, or scRNAseq). Illumina sequencing can also be adapted for other applications. For instance, proteins interact with DNA //in vivo// to form a dynamic structure called chromatin. Let's say you are interested in a DNA-binding protein called X. To find out what DNA sequences X binds to, you can extract and purify chromatin from cells, then use enzymes to gently cleave the DNA into small fragments under conditions in which X still binds to DNA. You can then purify X using antibodies and use Illumina sequencing to sequence the DNA fragments that co-purify with X. This procedure is called chromatin immunoprecipitation sequencing, or ChIPseq. There are many other similar applications too many to list and discuss here in detail. For small scale DNA sequencing, Sanger sequencing described above is still a commonly used method, although the cost for various NGS technologies have dropped so much that it is also starting to replace Sanger sequencing for small scale sequencing experiments. For instance, Nanopore sequencing (Fig. {{ref>Fig12}}) does not use DNA polymerase and can be used to replace Sanger sequencing in some routine applications. Nanopore sequencing also has the advantage of being able to detect DNA that has been chemically modified, such as methylated bases. This is relevant for studying things like epigenetic inheritance, and it something that neither Sanger nor Illumina sequencing can easily do.

{{ :different_ngs_tech.png?400 |}} Examples of next generation sequencing (NGS) technologies. Source: Bharti, R., and Grimm, D.G. Briefings in Bioinformatics 22(1), http://dx.doi.org/10.1093/bib/bbz155. Licensing: [[https://creativecommons.org/licenses/by/4.0/|CC BY 4.0]].

===== Questions and exercises ===== Exercise 1. How much tissue do you need for DNA sequencing? Each Sanger sequencing reaction uses about 1 μg DNA, and a typical Sanger reaction can give about 1 kbp of sequencing information before the reaction runs out. Let's assume that a typical human gene you want to study and sequence is 1 kb long. Usually to obtain DNA from human samples, we draw blood. To get enough human DNA to sequence this 1 kb gene, how much blood would you have to draw? You can assume the following: there are approximately 10,000 lymphocytes/mL of blood ; the average molecular weight of DNA is 330 Da/nucleotide (or 660 Da/base pair; 1 Da = 1 g/mol; 1 mol = 6x10²³ molecules); and the diploid human genome is approximately 6x10⁹ bp of DNA. We use lymphocytes (white blood cells) because red blood cells do not have nuclei (and therefore do not have DNA). Exact numbers are not super important when doing "back of the envelope" biology - what we care about here is the scale of the problem. Exercise 2: Draw three cycles of PCR by hand using colored pens and paper. It is very important to show the 5' and 3' ends of all the DNA molecules, and to keep track of which strands are template strands and which strands are newly synthesized in each cycle. Do not just copy from Fig. {{ref>Fig9}} - try to draw it based on your understanding of the reading, and then compare your drawing to Fig. {{ref>Fig9}}. What did you get right? What did you get wrong? Try again until you can get it fully correct.