Transcript Gaps - Nematode bioinformatics. Analysis tools and data

Genome Characterization
Assembly/sequencing
Assigned reading: Ch 9
BIO520 Bioinformatics
Jim Lund
The (original) genome sequencing
process
Organism Selection
Library Creation
Sequencing
Assembly
Gap Closure
Finishing
Annotation
The (current) genome sequencing
process
Organism Selection
Next gen. random sequencing lets
library generation get skipped
Sequencing
Assembly
Gap closure and finishing often
get skipped, at least for now.
Annotation
Contigs, Islands
contigs
Island
Assembly pipeline
1. Sequence reads.
2. Phred: base calling.
3. crossmatch: screen out vector, E.
coli sequence.
4. Phrap: assemble contigs.
5. Consed: view assembly, correct
problems.
6. Finishing.
Assembly Methods
•
•
•
•
Strip out vector (or contaminant)
Mask known repeats
Trim off unreliable data
Find Matches (n seq x n seq comparisons)
– how long (what ktuple [10 common])
– how perfect (reliability index)
– where to look? (ends only vs entire)
Assembly Programs
• PHRAP FAMILY
– phred/phrap/consed/cross_match
– Developed by Phil Green, U of Wash.
• Other assemblers
– phrap, kangaroo, phrapo,
– CAP, TIGRAssembler,...
http://www.phrap.org/
Assembly
• Phred -reads DNA sequencing trace files, calls bases,
and assigns a quality value to each called base.
– The quality value is a log-transformed error probability,
specifically: Q = -10 log10( Pe )
– Q = quality value, Pe = error probability.
– Q= 20 -> 1% chance of miscall, Q= 30 -> 0.1% chance of miscall.
• Phrap -assembles shotgun DNA sequence data.
• Consed/Autofinish -view, edit, and finish sequence
assemblies created with phrap.
– Allows the user to pick primers and templates
– Suggests additional sequencing reactions
– Suggest digests and forward/reverse pair information to
check accuracy of assembly.
Poisson statistics for
sequencing completion
L=read length
N=#reads
G=genome size
=e-L(N)/G
P0
Coverage
1 = 1-fold = 1X
1
% not
sequenced
37
3
5
8
0.03
10
0.005
50
< 1e20
E. coli 15kb
H. sapiens 900kb
Gaps
Number of Gaps = Ne-c
N=# of reads
c = fold coverage
150kb Target Clone, 500 bp reads
Coverage,
reads
1, 300
Gaps
5, 1500
10
8, 2400
1
10, 3000
0
50, 15000
0
111
Gaps
Number of Gaps = Ne-c
N=# of reads
c = fold coverage
Human genome, 3Gb, 1,000 bp reads
Coverage,
reads
1, 3e6
Gaps
1,000,000
5, 1.5e7
100,000
8, 2.4e7
8,000
10, 3e7
1,400
50, 3.75e8
7
454 Seq, 400bp reads
Contigs, Islands
C
T
T
T
contigs
Island
Finishing
• GOALS
– >95% coverage on BOTH strands
– every base covered 3X
– resolve ambiguities
• Finish when random no longer
productive (~8X range)
Sequence finishing.
How?
• Identify gaps, ambiguities
– Captured gaps: gaps is contained in a clone
• Extend from end of contigs
– Resequencing, new chemistry.
– Specific primers
– Subcloning and sequencing.
• Uncaptured gaps.
– New specific primers
– PCR across gap, sequence PCR product.
• Resolve ambiguities
– Consensus or resequence
• Specific primers, different chemistry
Large clone sequencing
process
• Phase 1: Unfinished, may be
unordered/unoriented contigs, with
gaps.
• Phase 2: Unfinished, fully oriented
and ordered sequence, may contain
gaps and low quality sequences
• Phase 3: Finished, no gaps.
Genome assembly after
initial contigs are made
• Order clones/contig sequences:
– Sequence overlaps.
• Clone/contig end sequences.
– Clone fingerprints.
– Anchor using other maps
• Sequence based markers on genetic or physical
maps.
• Conserved synteny to other genomes.
• Easiest when re-sequencing, e.g, another
human genome!
Process Control
• LIMS
– Laboratory
information
management
system
• AIMS
– Analysis
information
management
system
Hard genome sequencing
problems
• Repeats
• Complex genome structures
Where does a clone from a repetitive region map?
Approaches to sequence
repeat problems
• Multiple fragment sizes in 1 project
• Use length/distance info
• New assemblers, eg. ARACHNE
Results of Multi-length
Fragment Assembly
• Contigs
• “Supercontigs”
• Clone links for
finishing
• Clone map
DOE Joint Genome Institute (JGI) Prokaryote
Finishing Standards
• All low-quality areas (<Q30) are reviewed and
resequenced.
• The final error rate must be less than 0.2 per 10 Kb.
• No single-clone coverage is permitted (minimum of 2x
depth everywhere).
• Single-stranded regions are manually inspected and
quantified.
• All positions where an aligned high-quality read (>Q29)
disagrees with the consensus base are checked.
• All strings of xxxx are resolved in the final sequence.
• All repeats are verified.
• The ends of final contigs (chromosomes, plasmids) are
checked
• The final assembly is given a manual QC check.
Completed genomes
23 complete, 329 in assembly, in progress 389
Arabidopsis thaliana
Caenorhabditis elegans
Candida glabrata
Cryptococcus neoformans
Cyanidioschyzon merolae
Debaryomyces hansenii
Drosophila melanogaster
Encephalitozoon cuniculi
Entamoeba histolytica
Plants
Animal s
Eremothecium gossypii
Homo sapiens
Kluyveromyces lactis
Leishmzania major Friedlin
Mus musculus
Oryza sativa
Saccharomyces cerevisiae
Schizosaccharomyces pombe
Trypanosoma cruzi
Yarrowia lipolytica
Protists
Fungi
http://www.ncbi.nlm.nih.gov/genomes/leuks.cgi
Genomes Complete
• Eukaryotes--23 complete, 329 in assembly, in
progress 389
–
–
–
–
–
Human, mouse, rat, zebrafish,
Homo sapiens neanderthalensis
Drosophila, Anopheles, Caenorhabditis
Arabadopsis, oat, corn, barley, rice, tomato
Saccharomyces, Schizosaccharomyces,
Magnaportha, Cryptococcus, Candida…
– Encephalitozoon cuniculi, Guillardia theta
– Toxoplasma, Plasmodium
– And many more…
Eubacteria and Archaea genomes
• 608 Bacteria and 48 Archaea completed
• Comprehensive Microbial Resource
– http://pathema.tigr.org/tigrscripts/CMR/CmrHomePage.cgi
• Joint Genome Institute
– http://www.jgi.doe.gov/genome-projects/
– 2065 genome projects underway or
completed!
• NCBI Genomes
Genome Centers
• Joint Genome Institute (DOE)
• Whitehead Institute (MIT)
• TIGR
• Washington University (St. Louis)
• Celera
• Sanger Institute (the other UK)
• RIKEN (Japan)
• Beijing Genomics Institute (China)
• Max Planck (Germany)
…
Where do you find Genomic
data?
• NCBI
– Entrez (by clone, by Refseq)
– Genome (view and search map)
• Genome center sites
• Organism genome project sites
• Annotations projects
– UCSC Genome Browser,
– Ensembl Genome Browser
Arabidopsis
http://mips.helmholtz-muenchen.de/plant/athal/index.jsp
C. elegans (nematode)
http://wormbase.org