Today

1
Today

• Please read
• Science 291 1304-1315

2
Human Genome Project DissentersMy Brush with
Greatness?

• 1992 Two years into the HGP, two of the projects
  biggest critics were
• Sydney Brenner believed that the HGP should
  focus on human EST collections, and sequence the
  genome of a simple vertebrate (Fugu).
• Craig Venter believed that the clone-by-clone
  approach was not the most efficient way to
  proceed, suggested that shotgun approaches, and
  even a whole genome approach was feasible.

they were both right.
3
Sydney Brenner

• 2002 Nobel Prize (Medicine/Physiology)
• Sydney Brenner and John E. Sulston, Britain
• H. Robert Horvitz, United States
• for discoveries concerning how genes regulate
  organ development and a process of programmed
  cell death.

4
Expressed Sequence TagsESTs
Brenner was right.

• End sequenced cDNAs
• (complementary DNA)

• cDNA synthetic DNA transcribed from a mRNA
  template,
• through the action of an RNA dependant DNA
  polymerase called reverse transcriptase.

Online Primer est.html
5

• Still Sequencing cDNAs,
• first and easiest look into any genome,
• useful in understanding genomic sequence (gene
  finding),
• helps determine splice site variants,
• shorter than genomic clones, fits in plasmids,
• etc.

6
tissue specific ESTs are very useful.
7
Whole Genome Assembly
Venter was right.

• 1995 1.8 Mbp Haemophilus influenza genome
  sequenced,
• 1996 - on Mycoplasma, E. coli and others,
• 1999 Chromosome 2 of Arabidopsis,
• 2000 Drosophila (120 Mbp) genome,
• Human, Mosquito, etc
• Lots of genomes, several applications...

WGA of bacterial, viral populations...
8
(No Transcript)
9

• 1 year, 120 megabases,
• Assembly algorithms could generate accurate
  genomic sequences,
• Interim assemblies (or mapping) were not
  necessary.

24 MARCH 2000 VOL 287 SCIENCE
10
Big Biology
11
Think About This

• the plasmid library construction is the first
  critical step in shotgun sequencing,
• if the DNA libraries are not uniform in size,
  non-chimeric, and do not randomly represent the
  genome, then the subsequent steps cannot
  accurately reconstruct the genome sequence.
• We used automated high-throughput DNA sequencing
  and the computational infrastructure to enable
  efficient tracking of enormous amounts of
  sequence information (27.3 million sequence
  reads 14.9 billion bp of sequence).

12
Whos DNA?

• 21 enrolled donors,
• age, sex, ethnographic group,
• one African-American,
• one Asian-Chinese,
• one Hispanic-Mexican,
• two Caucasions.

13
Whos Mostly?
14
(No Transcript)
15
back to humans
Individuals, Libraries,
Sequence coverage, Clone coverage, Other?
What to know?
543 bp average sequence read
8, September 1999 - 25, June 2000
16
(No Transcript)
17
WGA Outline
18
DNA in sized libraries
19
back to humans
Individuals, Libraries,
Sequence coverage, Clone coverage, Other?
What to know?
543 bp average sequence read
8, September 1999 - 25, June 2000
20
Whole Genome Assembly

• 1. Screener
• 2. Overlapper
• 3. Unitigger/Discriminator,
• 4. Scaffolder,
• 5. Repeat Resolver.

21
Screener

• ...finds and masks microsatellite repeats,
  known repeated regions and ribosomal DNA,
• masked regions not used to make contigs,
• marks the rest for overlapping.

22
Overlapper

• ...looks for end-to end overlaps of at least 40
  bp with no more than 6 differences in match,

Whats the significance?
...a one in 1017 event.
given perfect randomness.
23
Good News

• ... uniquely assembled contigs (unitigs) are
  readily identifiable,
• all of the assembled sequences match over all of
  the known sequence,

- and -
...are consistent with an 8x sequence coverage.
24
Whole Genome Assembly

• 1. Screener
• 2. Overlapper
• 3. Unitigger/Discriminator,
• 4. Scaffolder,
• 5. Repeat Resolver.

25
Unitigs
But(t)
...the Screener doesnt include all of the low
frequency level repeats, ...so, a majority of
the Overlapper outputs turned out to be bogus.
26
What Now?

• over-collapsed assemblies are identified and
  broken down into unitigs when possible...
• these too-large contig sets are sent to the
  Unitigger/Discriminator.

27
Unitigger...differentiates between a true
overlap, and an overlap that includes more than
one loci.
28
Discriminator
29
Discriminator
...may yield u-unitigs.
Unitigger/Discriminator Output correctly
assembled contigs covering 73.6 of the genome.
30
Scaffolder

• ...contigs the contigs,
• uses mate-pair information, two or more
  consistent mate-pair matches yields 1 in 1010
  odds of being chance.

31
Repeat Resolver ...most of the remaining gaps
were due to repeats.
Rocks Use low Discriminator Value contig
sets to fill gaps, - find two or more mate
pairs with unambiguous matches in the scaffold
near the gap (2 kb, 10kb or 50 kb), (1 in
107),
Stones - find mate pair matches 2 kb, 10 kb,
and 50 kb from gap, place the mate in the gap,
check to see if its consistent with other
placed sequences.
32
Repeat Resolver ...most of the remaining gaps
were due to repeats.
Rocks Use low Discriminator Value contig
sets to fill gaps, - find two or more mate
pairs with unambiguous matches in the scaffold
near the gap (2 kb, 10kb or 50 kb), (1 in
107), Stones - find mate pair matches 2 kb,
10 kb, and 50 kb from gap, place the mate in the
gap, check to see if its consistent with other
placed sequences.
33
If that Doesnt Work

• ...find a mate-pair that spans the gap, and
  sequence it,

Chromosome Walking
34
Today/Friday

• Questions about WGA,
• CSA,
• Comparisons,
• Quality Control, etc.