Assembling and Annotating the Draft Human Genome

1
Tools for understanding the sequence, evolution,
and function of the human genome.
Jim Kent and the Genome Bioinformatics
Group University of California Santa Cruz
2
The Goal
Make the human genome understandable by humans.
3
Step 1
Sequence the human genome
4
Idealized Hierarchical Shotgun Sequencing
5
Mapping
300,000 BAC Clones Were Digested and Run on
Agarose Gels Cari Soderlunds FPC and Wash U
Pathfinders Made Fingerprint Map Contigs
Bob Waterston escaping management
Genetic and radiation hybrid maps placed contigs
on chromsomes
6
Sequence and Assembly

• BAC Clones shotgun sequenced at high throughput
  to 4x draft.
• Assembled with Phil Greens Phrap

7
GigAssembler
Jim Kent
David Haussler
(meanwhile Celera working on whole genome shotgun
version)
8
The Truth

• - ?
• ?
• - ? - -
• ? - ?

light
- darkness
Keeping strands straight is the hard part
9
Finishing Sequence

• Using primers to end of contigs close gaps.
• Checking automatic assembly especially near
  tandem repeats.
• Checking in-silico restriction digest of BAC
  matches actual digest.
• Time consuming - 1 year to draft genome, 2
  years to finish.
• Human finished. Mouse will be finished
  (currently half finished). Other genomes may
  stay at draft stage, though draft stage can be
  very good these days.

10
Now What?
TGGCTTTTGAAGGGAGTTCTGTTTATATATACGTCAACATCCAGTTGGAG
GTGAAAAGGTTAGCACTTGACCCAGGAAGTATCCATGTTTGTTTCAAAAA
TAAATCTGCTTCATAAATTTCTTCATCAGTCTTTTTTTCCATTATGAGC
TTTGATTATAATAAAGGAGCTGTTATTAACTTTTATTCAAGAAAAGGCCC
ATCTCTTTGAAAATATTTACCACCCTTCTCCCTTTCCCCTCATGAAATG
TGCCAACTTCATAGGAATTAACAAATTGTAGCCCAGCCAAATACACGGAT
GCTTAAGCATACCTGAAACTTGAGTATATTTATTTATTACAGACATCCT
AAGACCCGTAAACTCTGCTCTGGATCATATCACTCCAGGATCTCAGAGCT
GTTCATGATTGTACAGGAAATGGGGAATATCATAGGCTCACAAAGGATA
ACTGATAGAACTCAGTGTGGTACTTTGGGGACATCAAACATTGTGCGACA
TGCAAAAGACTATTCACGAATAACACAAAATATACATTCATTGTGCCAT
CCATCACATTAACAATTGAGCTGAAAATACATTATATCCAGCTAAGATAA
CTGTGGAAGGAAGAAATTGGTTTGAATAATACTTTTAGGTTCTGAATAA
CCCAGCACAAATTTTAAACAGAGGGTGGCCCGAGAAGAAAGGGGTAGAGA
TTGGGAAAGACTTAGCACAGGAAGCCGGGTTTCTGAAGTTTGTGCTCTG
CAGGGCTTCTTAACTGTAAGAACAAATCAAGGCTACCCTCTGAGGCATCT
GATTGGGTTTAAATGAGGGAATTTTTTCTTTCACCTATAAAATTGTACC
AGTTTAGAGAGTTTGCCCACCCTGTTTTAGTAACCTAAACATTTCTAGAA
AATCTGTATAAAGATAAATCTCTTAGGACAAAGTATTTACAACCAGCAA
ACTCACACACATGAAAATGACTTAAATTAAGGGATGAATTAATTGTGTAA
ACATATAGTGCATCTCTTCTTCCTGAGCTCCTGGACTCGCCTTTCGCTA
TATCCTACTTTCAAGGACAAGGGAGGGGAGAGCTGTACATATAGTTAGAT
AAAAGATGAGAAGATTCCTTCTGGCATGTTTCTGTTGGCAAAGGGAACT
ATTTTCCAAAAGGTCATCTGAAAGGAACAGTAGGTTCTGTGAATTCTCCT
AAAAGCAGGAGGGATGTTAAGGCCCACCAGAAAATGTATGCTGGCACCC
AATCTGGATGAAGGTGTTAACCCCGCACCAAGTCTCTGGTCCAGAATTAT
CTGCAAATATATTATCCTGGCCAGGAGCTCCCCAGATAGGATTAGAAAG
GAAGAAAGAGACTGTAAATGGAAAGAAAGATAAGCTAAGCATGTGCTTTG
GGTAAGAAGTCCCAGCCCAAGGAGATGCCTGGGCTGTTGTCTGGGGCTG
GAGCCGCCTCAGTGGGAGGTAGTCAGAGTGTCTGAGGTAGAAGACCCCGG
GGAAGGAACGCAGGGCGAAGAGCTGGACTTCTCTGAGGATTCCTCGGCC
TTCTCGTCGTTTCCTGGCGGGGTGGCCGGAGAGATGGGCAAGAGACCCTC
CTTCTCACGTTTCTTTTGCTTCATTCGGCGGTTCTGGAACCAGATCTTC
ACTTGGGTCTCGTTGAGCTGCAGGGATGCAGCGATCTCCACCCTGCGGGC
GCGCGTCAGGTACTTGTTGAAGTGGAACTCCTTCTCCAGTTCCGTGAGC
TGCTTGGTAGTGAAGTTGGTGCGCACCGCGTTGGGTTGACCCAGGTAGCC
GTACTCTCCAACTTTCCCTGGGGCAAAGTGGGAAGCCATGAGACGGAAA
TGTAAAAATTTTTAAATCGACTTGAGATTCCCCACACGCTTCATGGCAAC
ACTCAGGTAAAGAAAAGATCAAGAACTCAGCACAAATCGGGCTGTGGAG
GGTGAGTGATGAGGTGTAAAGTGTTAACCTGATGTAAACCATTAGCATGG
TCAGACCGGTGATTAATGGAGCCTCAAGATATTAACAGAACACTACCGT
CACAATAACCACCCCCACATACTTCCTATTTCCCAAATGTATAAAATCCT
TGAAAACACACCAATCCCTGAGACTTCTTTGCCCCAACACCTCTGGGCA
CCCTCTCCATGCACTACAACACTAGTCTGATACAAAAGCCTTTTAAAAAA
AAGATCATTATTAATTTCCTTGGAAATTAAGCATACCAGCTCCTTCCAG
AATAATCAAGGAGCATCCACCAACCAGCAGGACTGACCTGTTTTGGGAGG
GTTTCTTTTGACTTTCATCCAGTCAAAAGTCTGCGCTGGAGAAGATGTC
TCCGATGCGGGGGAGCGACAGGCTTCTTGGTGGCTGGCGTGGAGAGGGGA
CAAGGAGTTATTATACGTAGCCAGGGCCAGGCTCTGGTGCTCCTGTCCA
TATGAGTGGTGAATGTATTGAGGCGAGCCCACCGCGCCCCCAGCATAACC
CTGGTGGTGGTGGTGATGCTGGACCATGGGAGATGAGAGATTTCCAGAG
TAAACAGCGGGAGCGCACTGGGGGTACCCACCACTTACGTCTGCTTCCTG
ATTTAACGCGTAGGGGCTGTAAGGCGCACTGAAGTTCTGTGAGCCATAG
CTTGGACCACAACTTGAGTGGGAGTAGGACACCCCCAGGTTCCCGGAAGT
CTGGTAGGTAGCCGGCTGGGGGTGGCGATGGTGGTGGTGGTGGTGGTGG
TGGGGCGAACCGATCTGCACCCCCCTGCCCACTAGGAAGCGGTCGTCGCC
GCCGCAACTGTTGGCGCTGACCGCGCACGACTGGAAAGTTGTAATCCTA
TGGTCCGAGGGGTAGGCTCGGGCTGAGCAGGTCCCCGAGTCGCCACTGCT
AAGTATGGGGTATTCCAGGAAGGAGTTCATTCTTGCATTGTCCATCTGT
CACTGAGTGACCTGGTCCTGCGAAGCCCGGCGTGACTGTGCCAACTTTCT
CACTTCCTC
11
Finding the Genes
Dr. Blat helping a gene find itself.
12
SIGLEC7 - a gene with some transcriptional
complexity.
Sialic Acid Binding/Ig-like Lectin 7 displayed in
UCSC Genome Browser
13
Genes Lines of Evidence

• Full length human mRNA (the best!)
• Protein homology with other species.
• EST evidence - 1st step for much mRNA.
• Evidence from genome/genome alignments
• HMM based gene finders

14
Transferrin Receptor in UCSC Genome Browser
15
Transferrin
Clicking on a known gene brings up a large page
of information on the gene.
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
Current state of human genome

• 99 of human genome sequenced. Last 1 will
  still be a challenge.
• 85 of human genes located. Substantial
  resources are being devoted to last 15.
• 20 of human genes with any depth of functional
  annotation. Curation and integrated database are
  key to progress.
• lt1 of human regulatory regions located.

21
Transferrin Receptor
Note peaks of conservation in 3 UTR. These
include iron response elements which regulate
translation of this gene.
22
Comparative Genomics
Webb Miller
23
Comparative Genomics at BMP10
24
Conservation of Gene Features

• Conservation pattern across 3165 mappings of
  human RefSeq mRNAs to the genome. A program
  sampled 200 evenly spaced bases across 500 bases
  upstream of transcription, the 5 UTR, the first
  coding exon, introns, middle coding exons,
  introns, the 3 UTR and 500 bases after
  polyadenylatoin. There are peaks of conservation
  at the transition from one region to another.

25
Chaining Alignments

• Chaining bridges the gulf between syntenic blocks
  and base-by-base alignments.
• Local alignments tend to break at transposon
  insertions, inversions, duplications, etc.
• Global alignments tend to force non-homologous
  bases to align.
• Chaining is a rigorous way of joining together
  local alignments into larger structures.

26
Chains join together related local alignments
Protease Regulatory Subunit 3
27
Affine penalties are too harsh for long gaps
Log count of gaps vs. size of gaps in mouse/human
alignment correlated with sizes of transposon
relics. Affine gap scores model red/blue plots as
straight lines.
28
Before and After Chaining
29
Chaining Algorithm

• Input - blocks of gapless alignments from blastz
• Dynamic program based on the recurrence
  relationship score(Bi) max(score(Bj)
  match(Bi) - gap(Bi, Bj))
• Uses Millers KD-tree algorithm to minimize which
  parts of dynamic programming graph to traverse.
  Timing is O(N logN), where N is number of blocks
  (which is in hundreds of thousands)

jlti
30
Netting Alignments

• Commonly multiple mouse alignments can be found
  for a particular human region, particularly for
  coding regions.
• Net finds best match mouse match for each human
  region.
• Highest scoring chains are used first.
• Lower scoring chains fill in gaps within chains
  inducing a natural hierarchy.

31
Net Focuses on Ortholog
32
Net highlights rearrangements
A large gap in the top level of the net is filled
by an inversion containing two genes. Numerous
smaller gaps are filled in by local duplications
and processed pseudo-genes.
33
Useful in finding pseudogenes
Ensembl and Fgenesh automatic gene predictions
confounded by numerous processed pseudogenes.
Domain structure of resulting predicted protein
must be interesting!
34
Mouse/HumanRearrangement Statistics
Number of rearrangements of given type per
megabase excluding known transposons.
35
A Rearrangement Hot Spot
Rearrangements are not evenly distributed.
Roughly 5 of the genome is in hot spots of
rearrangements such as this one. This 350,000
base region is between two very long chains on
chromosome 7.
36
Reconstructed ancestral (boreutherian) genome for
one chromosome
37
Finding Function

• Weve located 85 of the genes, on track for 95
  in a year or two.
• We have SOME idea of what 30 of the genes do.
• We have virtually NO idea of what the rest do.

38
How to Find Function

• Homology - guilt by association. Orthologs very
  valuable.
• Genetics/knockouts - what happens when a gene
  gets broken?
• RNAi is speeding this up amazingly in worms and
  other model organisms.
• Expression - when and where is gene used?
• Microarrays, in situs, GFP fusions.
• Interactions - what molecules are touching?
• Yeast 2 hybrid, Immunoprecipitations
• Literature - finding out what we already know.

39
Data Mining
40
Gene Sorter - info on sets of genes
41
Sorted by homology
42
Sorted by genome distance
43
Coping with Bioinformatics Tower of Babel
44
(No Transcript)
45
Up in Testes, Down in Brain
46
Encode Project

• ENCyclopedia Of DNA Elements
• Pilot phase detailed experimental analysis of 1
  of genome in 40 different regions.
• Many types of experiments
• CHIP/CHIP
• DNAse hypersensitivity
• Tiling microarrays
• Deep comparative genomics
• Data available at genome.ucsc.edu via ENCODE link
  .

47
ENCODE Dnase I Hypersensitivity, CHIP/CHIP,
transcription data
48
ENCODE Dnase I Hypersensitivity, CHIP/CHIP,
transcription data
49
Close up of region
50
VisiGene

• Image browser for in-situ and other gene-
  oriented pictures
• Hopefully in the long run will have a million
  images covering almost all vertebrate genes.
• Currently has 6000 images covering 1000 mouse
  transcription factors courtesy of Paul Gray et al.

51
(No Transcript)
52
(No Transcript)
53
(No Transcript)
54
Gene Browser Staff

• Programming Hiram Clawson, Mark Diekhans,
  Rachel Harte, Angie Hinrichs, Fan Hsu, Andy Pohl,
  Kate Rosenbloom, Chuck Sugnet,
• Docs, quality, support Gill Barber, Ron Chao,
  Jennifer Jackson, Donna Karolchik, Bob Kuhn,
  Crystal Lynch, Ali Sultan-Qurraie, Heather
  Trumbower
• Computer systems Jorge Garcia, Patrick Gavin,
  Paul Tatarsky

55
Comparative Genomics

• UCSC - Robert Baertsch, Gill Bejerano, Yontoa Lu,
  Jacob Pedersen, Katie Pollard, Adam Siepel, Daryl
  Thomas, David Haussler
• PSU - Laura Elnitski, Belinda Giardine, Ross
  Hardison, Minmei Hou, Scott Schwartz, Webb
  Miller,

56
Data Contributors

• Human Genome Project
• Genbank/DDJ/EMBL contributors
• Novartis GNF foundation
• Affymetrix, Perlegen, SNP Consortium
• SwissProt, Ensembl, EBI and NCBI
• Jackson Labs, RGD, Wormbase, Flybase
• Many contributors of gene prediction and other
  tracks.

57
Funding

• National Human Genome Research Institute
• Howard Hughes Medical Institute
• Taxpayers in the USA and California

58
THE END
59
Confounded Pseudogenes!

• Pseudogenes confound HMM and homology based gene
  prediction.
• Processed pseudogenes can be identified by
• Lack of introns (but 20 of real genes lack
  introns)
• Not being the best place in genome an mRNA aligns
  (be careful not to filter out real paralogs)
• Being inserted from another chromosome since
  dog/human common ancestor (breaking synteny).
• High rate of mutation (Ka/Ks ratio).
• Robert Baertsch at UCSC has produced a processed
  pseudogene track.
• Yontoa Lu working on a non-processed pseudogene
  track.

60
Close up of two processed pseudogenes
61
Detail Near Translation Start
Note the relatively conserved base 3 before
translation Start (constrained to be a G or an A
by the Kozak Consensus sequence, and the first
three translated bases (ATG).
62
Normalized eScores
63
Table browser - text-oriented browsing and data
analysis of genome browser database.