Transcript Transcriptome - Nematode bioinformatics. Analysis tools and data

The Transcriptome
Gene Discovery
Quantitation of Gene Expression
Reading: Ch 15.1
BIO520 Bioinformatics
Jim Lund
WHY?
• The genes (proteins) expressed determine the
state of the cell.
–
–
–
–
Signaling.
Metabolic capabilities.
Differentiation state (cell type).
Response to changes in environment.
• Verifies gene predictions.
• Transcriptional regulation
– Normal vs. abnormal
– Conditional expression
Transcriptome Analysis
• Gene (transcript) discovery
– transcripts
– alternative splicing/processing
•
•
•
•
Transcript assays
Promoter analysis
Transcription Factors
Cellular control networks
Gene Discovery
• Inference from genomic DNA
–Prokaryotes & fungi OK
• cDNA characterization
–EST
–SAGE
EST
(Expressed Sequence Tag)
• Sequence cDNA libraries
–proportional libraries
–subtracted or normalized
libraries
• Which end?
–5’ or 3’ or Whole
Library Type
• “regular” or
proportional
• Subtracted
– Miss
alternate
transcripts
• normalized
• Tissue
• Primer
– dT vs random
Ideal cDNAs
“Real” cDNAs
Which end?
• Whole cDNA
– BEST & HARDEST (Long)
• 3’-end
– Consistent technically, limited information
• 5’end
– Coding “identity” highest
• 5’ AND 3’
– Good, but technical & informatic challenge
Gene Discovery-Yeast
1
Fraction Known
0.8
0.6
0.4
0.2
0
# of ESTs (log scale!)
EST Data Analyses
• Clustering Analysis
–
–
–
–
Assemble ESTs into genes.
Alternative splicing forms
Find coding SNPs.
Truncated, unspliced, and junk ESTs can be
misleading
– Project: Unigene
– Program: stackPACK
• Frequency analysis
– Digital Differential Display
• DDD is a computational method for comparing
sequence-based gene representation profiles
among individual cDNA libraries or pools of
libraries.
EST Results (old)
• Known genes (30%)
• Similarities to other ORFs, ESTs
(30%)
– Infer Function?
• Novel Class (30%,  w/ time)
Typical Progress/Results
• Humans
– 6,694,833 ESTs
– 124,179 clusters (“sets”)
• 29,000 sets contain EST and mRNA seqs.
– CGAP EST library ”plateau” broken by:
• different tissues, different states
• normalized libraries
Data Quality
Considerations
• 99% correct data (1% errors!).
• Frameshifts-effects depend on tools
– BLASTX tool to “find” frameshifts
• How sensitive?
– TBLASTX, TBLASTN to “use” in other
projects
• How sensitive?
Gene Expression Assays
• EST (Poor method)
• SAGE
• Microarray Hybridization
• Next Gen Sequencing.
• Transcriptional Fusions
– GFP, LacZ fusions
Serial Analysis of Gene Expression
(SAGE)
1. Collect mRNA
2. Isolate short oligomers from each
transcript.
3. Ligate together the oligomers and clone
them.
4. Sequence thousands of clones.
5. Map the 1x104 – 1x105 oligomers to their
genes.
6. Find which genes are transcribed and their
relative expression levels.
7. http://www.sagenet.org (Vogelstein at JHU)
SAGE technique
• Prepare
biotin
labeled
cDNA
• Cleave
with
anchoring
enzyme
(NlaIII)
SAGE technique
• Ligate on
linkers
• Cleave
with
tagging
enzyme
(BsmFI)
SAGE technique
• Ligate, PCR,
and gel purify
ditags (102bp).
• Recleave with
anchoring
enzyme (NlaIII),
ligate to form
concatemers.
• Size select,
clone and
sequence
concatemers.
Colon cancer vs. normal colon epithelium (SAGE)
Microarray Hybridization
• Determine gene expression by
parallel hybridization of labeled
cDNA to DNA attached to a fixed
support.
• http://cmgm.stanford.edu/pbrown/
Microarray Hybridization
• Producing chips
• Producing probes / reading arrays
• Analyzing and interpreting data
Transcriptional Array
orf 1 orf 2 orf 3
1
2
3
4
5
6
7
8
9
3 cm
200 spots
2
Condition 1
mRNA
Condition 2
mRNA
40,000 dot/9 cm
or
> All human genes
Transcriptional Array-1
orf 1 orf 2 orf 3
1
22
3
4
5
6
7
88
9
3 cm
200 spots
2
Condition 1
mRNA
Condition 2
mRNA
40,000 dot/9 cm
or
> All human genes
Transcriptional Array-2
orf 1 orf 2 orf 3
1
22
3
4
5
6
7
88
9
3 cm
200 spots
2
Condition 1
mRNA
Condition 2
mRNA
40,000 dot/9 cm
or
> All human genes
Microarray Technologies
• Spotted arrays (Brown et al.)
– Spot arrays on glass slides
– PCR fragments
– Long (50-70bp) oligo arrays
• Synthesis
– Affymetrix
(www.affymetrix.com)
• High density array of 25 bp oligos
• Made using light directed oligonucleotide
synthesis and photolithography
– Agilent, CombiMatrix
• Made using light directed oligonucleotide
synthesis and mirrors.
Spotted Arrays
Print Quill
Spotted microarray image
Affymetrix photolithographic technology
•Lithographic masks are used to either block or transmit light onto specific locations of the array.
•The surface is then flooded with a solution containing either adenine, thymine, cytosine, or guanine,
and coupling occurs only in those regions on the glass that have been deprotected through
illumination.
•The coupled nucleotide also bears a light-sensitive protecting group, so the cycle can be repeated.
•Microarray is built as the probes are synthesized through repeated cycles of deprotection and
coupling.
•Typically ends at 25 bps.)
•Current arrays have 1.3 million unique features per array.
GeneChip Expression
Assay Design
Affymetrix GeneChips:
Expression Analysis
• Available for humans and model
organisms.
• Made only by Affymetrix.
• Chip designs change slowly.
• GeneChips:
– Human: 50,000 RefSeq genes and ESTs
– C. elegans: 22,500 genes (12/00 genome
annotation)
– Rat 230: 30,000 genes, ESTs
– Yeast: 6100 gene set
– Tiling arrays for model organisms
• http://affymetrix.com
Quantitation of fluorescence signals
(Image to data)
1. Hybridization, scan in chip image.
2. Gridding
– Determine where the spots are.
3. Spot intensity and local background
determination.
4. Normalization
– Adjust to make the red and green total signal
intensities the same.
5. Gene expression ratio.
– Red channel/green channel.
•
Programs:
– ScanAlyze, http://rana.lbl.gov/EisenSoftware.htm
– GenePix,
http://www.moleculardevices.com/pages/instruments/mic
roarray_main.html
Microarray data
Big tables of numbers!
Viewing microarray data
Scatter plot: log(ch1) vs log(ch2)
M vs A:
expression levell vs expression change
Clustergram
Volcano plot: log(expr) vs p-value