mapping_tutorial_interactive

Report
Getting the computer setup
• Follow directions on handout to login to
server.
• Type “qsub -I” to get a compute node.
• The data you will be using is stored in
../shared/
Mapping RNA-seq data
Matthew Young
Alicia Oshlack
Bernie Pope
Illumina Sequencing Technology
3’ 5’
DNA
(0.1-1.0 ug)
A
G
C
T
G
C
T
A
C
G
A
T
A
C
C
C
G
A
T
C
G
A
T
A
T
C
G
A
T
G
C
T
Sample
preparation
Single
molecule
Cluster
growtharray
5’
Sequencing
1
2
3
4
5
6
7
8
9
T G C T A C G A T …
Image acquisition
Base calling
Slide courtesy of G Schroth, Illumina
Raw data
• Short sequence reads
• Quality scores = -10log10(p) or similar…
@SEQ_ID
GATTTGGGGTTCAAAGCAGTATCGATCAAATAGTAAATCCATTTGTTCAACTCACAGTTT
+
!''*((((***+))%%%++)(%%%%).1***-+*''))**55CCF>>>>>>CCCCCCC65
Handout Reference: Page 2
Quality checks
• Base composition
• Quality score
• PCR Artifacts
Sequencing transcripts not the genome
Reads
Gene
CDS
CDS
CDS
CDS
transcript
CDS
CDS
CDS
CDS
Difficulty here is that reads spanning a exon-exon junction may not get mapped when
mapping to the genome.
One strategy: Supplement the reference genome with sequences that span all known or
possible junctions.
CDS
Coding Sequence
CDS
Exons
Introns
Aggregate reads based on:
 exons
 Exons + junctions
 All reads start to end of transcript
 De novo methods
CDS
CDS
Splice Junctions
Mapping tools
Type
Name
Link
General aligner
GMAP/GSNAP
http://research-pub.gene.com/gmap/
BFAST
http://sourceforge.net/apps/mediawiki/bfast/index.php
BOWTIE
http://bowtie-bio.sourceforge.net/index.shtml
CloudBurst
http://sourceforge.net/apps/mediawiki/cloudburst-bio/index.php
GNUmap
http://dna.cs.byu.edu/gnumap/index.shtml
MAQ/BWA
http://maq.sourceforge.net/
Perm
http://code.google.com/p/perm/
RazerS
http://www.seqan.de/projects/razers.html
Mrfast/mrsfast
http://mrfast.sourceforge.net/manual.html
SOAP/SOAP2
http://soap.genomics.org.cn/
SHRiMP
http://compbio.cs.toronto.edu/shrimp/
QPALMA/GenomeMapper/PALMapper
http://www.fml.tuebingen.mpg.de/raetsch/suppl/palmapper
SpliceMap
http://www.stanford.edu/group/wonglab/SpliceMap/
SOAPals
http://soap.genomics.org.cn/
G-Mo. R-Se
http://www.genoscope.cns.fr/externe/gmorse/
TopHat
http://tophat.cbcb.umd.edu/
SplitSeek
http://solidsoftwaretools.com/gf/project/splitseek
Oases
http://www.ebi.ac.uk/~zerbino/oases/
MIRA
http://sourceforge.net/apps/mediawiki/mira-assembler/index.php
De Novo annotator
De Novo transcript
assembler
Getting the computer setup
• Follow directions on handout to login to
server.
• Type “qsub -I” to get a compute node.
• The data you will be using is stored in
../shared/
Familiarizing yourself with bowtie
• The minimum information bowtie needs is
your reads (i.e., the fastq file from the
machine) and a reference.
• The reference is the genome or transcriptome
we are trying to map to, transformed using a
Burrows Wheeler Transform to allow fast
searching.
• Many optional parameters to tweak
alignment.
Handout Reference: Pages 2-7
How do you map 10^9, 76bp
sequences, to a 10^9 bp reference
• Ideally we’d test every position on the
genome for its suitability as a match and
assign it a score based on # mismatches,
indels etc.
• However, this is computationally impossible,
so we have to come up with something else.
Handout Reference: Pages 2-7
Lots of aligners, one general strategy
• Quick “heuristic” is performed to cut down
the number of candidate alignment regions
for each read.
• More precise algorithm is employed to decide
which of these candidates is a valid alignment.
Handout Reference: Pages 2-7
A fragment as seen by an aligner
;
Seed
Read
Fragment
Burrows Wheeler acts on this
Smith-Waterman acts on this
• Heuristic acts only on the seed. Putative mapping
locations identified.
• A more precise algorithm extends each seed to
the full read and ranks them.
• The fragment is not sequenced directly, it’s
sequence is inferred based on mapping of the
read.
Handout Reference: Pages 2-7
Our test data set
• 76bp single end reads.
• Sequenced using the Illumina Genome
Analyzer
• RNA taken from Mouse myoblast cell line.
• We only look at a subset of one lane, full data
available here
http://www.ncbi.nlm.nih.gov/geo/query/acc.c
gi?acc=GSE2084
Getting the computer setup
• Follow directions on handout to login to
server.
• Type “qsub -I” to get a compute node.
• The data you will be using is stored in
../shared/
A strict mapping
bowtie -t -p 1 -n 0 -e 1 --sam --best --un
Strict_Unmapped.fastq ../shared/BowtieIndexes/mm9
../shared/Sample_reads.fastq Strict_Mapped.sam
• -n sets the number of mismatches in the seed
• The sum of quality scores at ALL mismatches (not just the
seed) must be less than -e.
• --un saves the unmapped reads to the file specified
• -t prints timing information
• -p sets the number of simultaneous threads
• --sam makes bowtie output in SAM format
• --best ensures the best map is returned
Handout Reference: Page 8
Using the defaults
bowtie -t -p 1 --best --sam
../shared/BowtieIndexes/mm9
../shared/Sample_reads.fastq test.sam
•
•
•
•
•
-t prints timing information
-p sets the number of simultaneous threads
--sam makes bowtie output in SAM format
--best ensures the best map is returned
--un saves the unmapped reads to the file
specified
Handout Reference: Page 8
Allowing some mismatches
Bowtie -t -p 1 -n 3 -e 200 --sam –best --un
Loose_Unmapped.fastq
../shared/BowtieIndexes/mm9
Strict_Unmapped.fastq Loose_Mapped.sam
• Set the number of seed mismatches to the
maximum.
• Set -e to a value more appropriate for our read
length.
• How many reads do you map?
Handout Reference: Pages 8-9
A closer look at our data set: fastQC
Handout Reference: Pages 9-10
Handout Reference: Pages 9-10
Trimming reads
• Bowtie allows you to trim reads before it
attempts to map them.
• You can trim from the left (5’) end of the read
with the --trim5 option.
• You can trim from the right (3’) end of the
read with the --trim3 option.
Handout Reference: Page 10
Sequencing transcripts not the genome
Reads
Gene
CDS
CDS
CDS
CDS
transcript
CDS
CDS
CDS
CDS
A library containing these bits of sequence (which do not appear in the genome) can
help map junction reads. This is called an exon-junction library.
A reference built from this library is in BowtieIndexes, called
mm9.UCSC.knownGene.junctions (named for the annotation it was built from).
Handout Reference: Pages 10-12
Options to map more reads…
• Trim some bases from the end of the reads
using --trim5 and/or --trim3.
• Map to the junction library instead of the
mouse genome using the
mm9.UCSC.knownGene.junctions index.
Handout Reference: Pages 9-12
Trim
[[email protected] Starting]$ bowtie -t -p 1 -n 2 -e 70 --trim5 15 -trim3 25 --sam --best --un Trimmed_Unmapped.fastq
../shared/BowtieIndexes/mm9 Loose_Unmapped.fastq
Trimmed_Mapped.sam
Time loading forward index: 00:00:02
Time loading mirror index: 00:00:02
Seeded quality full-index search: 00:00:47
# reads processed: 531414
# reads with at least one reported alignment: 164256 (30.91%)
# reads that failed to align: 367158 (69.09%)
Reported 164256 alignments to 1 output stream(s)
Time searching: 00:00:53
Overall time: 00:00:54
Handout Reference: Page 10
Then map to junctions
[[email protected] Starting]$ bowtie -t -p 1 -n 3 -e 200 --sam --best
--un Junctions_Unampped.fastq
../shared/BowtieIndexes/mm9.UCSC.knownGene.junctions
Trimmed_Unmapped.fastq Junctions_Mapped.sam
Time loading forward index: 00:00:35
Time loading mirror index: 00:00:34
Seeded quality full-index search: 00:00:24
# reads processed: 367158
# reads with at least one reported alignment: 128756 (35.07%)
# reads that failed to align: 238402 (64.93%)
Reported 128756 alignments to 1 output stream(s)
Time searching: 00:01:43
Overall time: 00:01:45
Handout Reference: Page 11
Number of mapped reads
Mapping
strategy
Command
line options
No. Mapped Reads
No. Unmapped
Reads
Reference
Strict
-n 0 -e 1
1,049,050 (47.32%)
1,167,992 (52.68%)
Genome
Loose
-n 3 -e 200
1,783,048 (80.42%)
433,994 (19.58%)
Genome
Trimming
-n 3 --trim5
15 --trim3 25
1,912,003 (86.24%)
305,039 (13.76%)
Genome
Junctions
-n 3 -e 200
2,007,627 (90.55%)
209,415 (9.45%)
Junction
Library
Handout Reference: Page 12
Further options
Type
Name
Link
General aligner
GMAP/GSNAP
http://research-pub.gene.com/gmap/
BFAST
http://sourceforge.net/apps/mediawiki/bfast/index.php
BOWTIE
http://bowtie-bio.sourceforge.net/index.shtml
CloudBurst
http://sourceforge.net/apps/mediawiki/cloudburst-bio/index.php
GNUmap
http://dna.cs.byu.edu/gnumap/index.shtml
MAQ/BWA
http://maq.sourceforge.net/
Perm
http://code.google.com/p/perm/
RazerS
http://www.seqan.de/projects/razers.html
Mrfast/mrsfast
http://mrfast.sourceforge.net/manual.html
SOAP/SOAP2
http://soap.genomics.org.cn/
SHRiMP
http://compbio.cs.toronto.edu/shrimp/
QPALMA/GenomeMapper/PALMapper
http://www.fml.tuebingen.mpg.de/raetsch/suppl/palmapper
SpliceMap
http://www.stanford.edu/group/wonglab/SpliceMap/
SOAPals
http://soap.genomics.org.cn/
G-Mo. R-Se
http://www.genoscope.cns.fr/externe/gmorse/
TopHat
http://tophat.cbcb.umd.edu/
SplitSeek
http://solidsoftwaretools.com/gf/project/splitseek
Oases
http://www.ebi.ac.uk/~zerbino/oases/
MIRA
http://sourceforge.net/apps/mediawiki/mira-assembler/index.php
De Novo annotator
De Novo transcript
assembler
Handout Reference: Pages 12-13

similar documents