The Past, Present, and Future of DNA Sequencing

Craig A. Praul Co- Director Genomics Core Facility Huck Institutes of the Life Sciences Penn State University

A very short history of DNA sequencing

I started from the conviction that, if different DNA species exhibited different biological activities, there should also exist chemically demonstrable differences between deoxyribonucleic acids.

Edwin Chargaff

Milestones

• • • • • • First Isolation of DNA : 1867 (Freidrich Meisher) Composition of nucleic acids; tetranucleotide theory : 1909 - 1940 (Phoebus Levine) G=C and A=T however, the G/C and A/T content of different organisms vary : 1950 (Edwin Chargaff) G/C content measured by annealing : 1968 (Mandel and Marmur) Maxam-Gilbert and Sanger Sequencing : 1977 Next-Generation Sequencing : 2005

Genomes Sequenced

• Virus – 3222 (Bacteriophage phi X 174, 5386 nt – 1977) • Bacteria – 2289 (Haemophilus influenza, 1.8 x 10 6 nt – 1995) • Eukarya – 168 (S. cerevisiae 1.2 x 10 7 nt – 1995; H. sapien, 3 x 10 9 nt -2001) • Archaea – 152 (Methanococcus jannaschi , 1.7 x 10 6 nt – 1996)

Next-Generation Sequencing

Liu et al. Journal of Biomedicine and Biotechnology Volume 2012 (2012), Article ID 251364, 11 pages doi:10.1155/2012/251364

Changes in instrument capacity*

ER Mardis. Nature 470 , 198-203 (2011) doi:10.1038/nature09796

Sequencing Cost

Date Sep-01 Sep-02 Oct-03 Oct-04 Oct-05 Oct-06 Oct-07 Oct-08 Oct-09 Oct-10 Oct-11 Oct-12 Jan-13 Cost per Mb $5,292.39

$3,413.80

$2,230.98

$1,028.85

$766.73

$581.92

$397.09

$3.81

$0.78

$0.32

$0.09

$0.07

$0.06

Cost per Genome $95,263,072 $61,448,422 $40,157,554 $18,519,312 $13,801,124 $10,474,556 $7,147,571 $342,502 $70,333 $29,092 $7,743 $6,618 $5,671 Source - NHGRI : http://www.genome.gov/sequencingcosts/

Central Dogma of Molecular Biology James Watson version - 1965

DNA RNA

So once we have the genomic DNA sequence of a species we have all of the information there is?

Really?

Protein

•

No, not really.

Illumina HiSeq and MiSeq

• Massively parallel – HiSeq : 150 or 180 million reads per lane – MiSeq : 15 million reads per run • Intermediate Read Length – HiSeq : 100 nt or 150 nt – MiSeq : 250 nt • High total output per run – HiSeq : 90 GB or 288 GB – MiSeq : 8 GB

Single Read Paired-end read

Sequencing Types

Mate-pair read

Library Types

• Many different library preps : DNA, mate-pair, mRNA, miRNA, ChIP • Fragmentation – DNA : 300 – 500 nt – RNA : 150 – 200 nt • Attachment of appropriate adapters – Complex : flow cell binding, F & R sequencing, BC – Custom : Avoid if possible • Removal of dimers/small inserts • Amplification (or not)

Applications

• de Novo sequencing (genomes, transcriptomes) • Resequencing (genomes, exomes, custom sequence capture) • RNA-seq (mRNA, miRNA, degradome) • Chip-Seq • Methyl-seq • RIP-seq • Amplicon

de Novo Experimental Design

• Estimate of genome size • Coverage (30 x – 100 x) • Sequencing Type (paired-end or mate-pair) • Example 100 MB genome, 100 x 100 nt paired-end reads – (100 MB) x (30 x coverage) = 3 GB – 3 GB / (200 nt for each pair of paired-end reads) = 15 million read pairs • Replicates

Resequencing : Sequence Capture

RNA-seq Experimental Design

• Estimate of transcriptome size (1-5% of genome ?) • Coverage (30 x ?) – mRNA or rRNA depleted RNA – Relative abundance of transcripts you are interested in • Sequencing Type (single read or paired-end) – Simple transcriptome vs. complex transcriptome – Splice variants • Example 3 GB genome, 100 nt single reads – (3 GB genome) x ( 5% transcriptome ) = 120 MB Transcriptome – (120 MB transcriptome) x (30 x coverage) = 4.5 GB total sequence – 4.5 GB / (100 nt for each read) = 45 million read pairs • Replicates : Yes!!!!

– Biological not technical

ChIP-Seq

http://www.nature.com/nmeth/journal/v4/n8/images/nmeth0807-613-F1.gif

RIP-seq Source : http://openi.nlm.nih.gov/imgs/rescaled512/3269675_ijms-13-00097f6.png

Methyl-seq 20 different types of base modifications in DNA are known and there are perhaps 200 modifications of RNA

Experimental Space: Next-Gen Platform

• PacBio : 0.075 x 10 6 – Whole transcript reads/sample, 1000 – 3000 nt • Roche 454 FLX+ : 0.5 -1 x 10 6 – reads/sample, 800 -1000 nt Small – Medium Genome de novo sequencing – Long Amplicon – Transcriptome • PGM: 1-2 x 10 6 – reads per sample, 400 nt Small genome de novo – Medium Amplicon • MiSeq: 1-2 x 10 6 – reads per sample, 50 – 250 nt Small genome de Novo – Small Amplicon • HiSeq : 10-100 x 10 6 – reads per sample, 50 – 150 nt Counting Applications : RNA-seq, ChIP-seq, RIP-seq, Methyl-seq – Large genome de novo and resequencing

Experimental Space: The Relevancy of “Classic” Techniques

Differential Gene Expression • Northern blotting (1977) : 1 Probe – 20 samples • Dot Blots (1987) : 100s of probes – 1 sample • RT-PCR (1992) : 100s of probes – 10 -100 samples • Microarrays (1995 ) : 100,000s of probes – 1 sample • Next-gen sequencing (2005) : 10-100 x 10 6 reads – 1 sample

The Future • More Reads • Longer Reads • Faster Sequencing • Cheaper Sequencing • New Applications