HighThroughput Oligo Design and Microarray Data Analysis

1
High-Throughput Oligo Design and Microarray Data
Analysis
Xiaowei Wang Department of Molecular
Biology Massachusetts General Hospital, Boston
2
Talk Summary

• Real-Time PCR Primer Design
• siRNA Design
• Microarray Data Analyses For NF-?B Pathway
• Mammalian Gene Association Network

3
Primer Design Problems

• Current primer design programs do not consider
  primer specificity.
• Researchers need to design their own primers.
• 60 primer design success rate using SYBR
  Green.
• Need to redesign primers, wasting time and money.

4
Why Using PrimerBank?
User comment on PrimerBank I demonstrated
accessing your database during a workshop I
think the audience got the message about which
software tool they should use to design assays
... i.e., none at all

• Benefits
• Pre-designed primers for 40,000 human and mouse
  genes
• 99 design success rate (validated over 1000
  primers)
• Free to the public

5
Primer Design Criteria
Specificity No repetitive 15-mer low BLAST
score highly unique 3 end reasonable sequence
complexity, etc. Sensitivity No secondary
structure no primer dimer, etc. Uniformity Simi
lar Tm, GC content, primer length, amplicon
length, etc.
6
Real-Time PCR of cytochrome P450 genes
b) Melting Curves
a) Amplification Plots
7
Gel Electrophoresis of The PCR Products
1 2 3 4 5 6 7 8
9 10 11 12 13 14 15 16 17
500 bp
350 bp
275 bp
225 bp
175 bp
125 bp
8
PrimerBank Home Page
http//pga.mgh.harvard.edu/primerbank/
9
(No Transcript)
10
Statistics for PrimerBank Usage
Total queries 34,201 Unique Users 2,332
11
Talk Summary

• Real-Time PCR Primer Design
• siRNA Design
• Microarray Data Analyses For NF-?B Pathway
• Mammalian Gene Association Network

12
qPCR Primer Design Criteria
Specificity No repetitive 15-mer low BLAST
score highly unique 3 end reasonable sequence
complexity, etc. Sensitivity No secondary
structure no primer dimer, etc. Uniformity Simi
lar Tm, GC content, primer length, amplicon
length, etc.
13
siRNA Design Criteria
Specificity No repetitive 15-mer low BLAST
score reasonable sequence complexity,
etc. Sensitivity No secondary
structure Uniformity siRNA length GC content
etc.
14
Challenges In siRNA Design

• RNAi mechanism is complex many factors could
  affect siRNA efficiency
• No well-defined guideline for siRNA efficiency
• Need to absorb the latest progress on RNAi
  studies

15
New Findings on siRNA Efficiency
Asymmetry in the assembly of the RNAi enzyme
complex. Schwarz DS, Hutvagner G, Du T, Xu Z,
Aronin N, Zamore PD. Cell. 2003 Oct
17115(2)199-208. Functional siRNAs and miRNAs
exhibit strand bias. Khvorova A, Reynolds A,
Jayasena SD. Cell. 2003 Oct 17115(2)209-16.
16
Functional siRNA Has Unstable 5-End
Functional siRNAs Non-functional siRNAs
Cell 115 209 (2003)
17
Cell 2003 115 (2) 199
18
Talk Summary

• Real-Time PCR Primer Design
• siRNA Design
• Microarray Data Analyses For NF-?B Pathway
• Mammalian Gene Association Network

19
NF-?B Functions

• Inflammatory and stress responses
• Apoptosis
• Cell proliferation and differentiation

20
NF-?B Signaling Pathway
Nature Rev Immunol (2002) 2725
21
?B-ras1 Binds and Stabilizes I?B
?B-Ras1
I?B-ß
PEST domain
22
Microarray Experimental Design
WT1 KO1 WT2 KO2
WT3 KO3
slide1 slide2 slide3 slide4
slide5 slide6
23
Microarray Data Processing Pipeline
Visually inspect slide images for bad spots
Filter out insignificant spots or high background
Identify statistically significant genes across
replicated slides
Generate a gene list sorted by significance
24
Global Gene Expression Changes
Sample Signals
Control Signals
25
The Most Significantly Down-regulated Genes In
Knockout Mice
Gene Ratio P-value Description
26
Cytochrome P-450

• Broad biological functions
• blood homeostasis
• cholesterol and steroid biosyntheses
• drug metabolism
• Diseases involved
• hypertension, diabetes, obesity, etc.

27
Conclusions

• ?B-ras1 down-regulates cytochrome P450 gene
  expression.
• 2. ?B-ras1 may regulate lipid metabolism in
  liver cells.

28
Gene Expression Regulation by NF-?B
Transcription Starting Site
NF-?B
?B site
NF-?B regulated gene
29
Analysis on Genes with ?B Sites
All arrayed genes
NF-?B activation
Microarray
?BSearch
Significant gene expression changes
Genes with ?B sites
Combined analysis
30
Differentially Expressed Genes With ?B Sites
Gene Ratio P-value Description
31
Talk Summary

• Real-Time PCR Primer Design
• siRNA Design
• Microarray Data Analyses For NF-?B Pathway
• Mammalian Gene Association Network

32
Gene Coexpression

• Microarray expression profiling can be used to
  identify coexpressed genes
• Gene coexpressions infer functional gene
  associations

33
Data Sources

• Microarray data NCBI GEO database
• Orthologous gene map NCBI HomoloGene
• Functional categories Gene Ontology Consortium

34
Strategy Outline
Microarray Data
Human-mouse gene map
Conserved gene coexpressions
Functional categories
Functional associations
Predict gene functions
35
High-Throughput Microarray Expression Profiling

• Affymetrix data from 69 experiments
• 746 human arrays
• 847 mouse arrays

36
Microarray Data Analysis
Normalize all 1593 Affy chips
Identify the expression profile for every gene
Calculate coexpressions for all combinations of
two genes
Determine the significance (P values) of the
coexpressions
Correct P values by False Discovery Rate
Coexpression clusters
37
NF- ?B Coexpression Clusters
38
NF- ?B Coexpressed Genes
39
Functional Association Scores
For each coexpression cluster
Ph and Pm - Coexpression p values for human and
mouse, respectively Sgo number of genes in one
functional category
40
Functional Associations
Functional Association Scores
Monte Carlo simulations
False Discovery Rate
Significant functional associations
Predict gene functions
41
Predict Gene Functional Associations

• Gene APPBP1
• Associated with Alzheimer's disease
• Function E1 ubiquitin-activating enzyme

42
APPBP1 Functional Associations
43
Predict Gene Functional Associations

• Gene HM9088
• Gene mutation leads to Parkinsons disease
• Function unknown

44
HM9088 Functional Associations
45
Acknowledgments

• Brian Seed
• Seed Lab
• Nancy Spandidos, Yi Yang
• Freeman Lab
• Mason Freeman, Harry Bjorkbacka
• The DNA Core Facility
• Glenn Short, Shukui Guan, Tao Lan
• Bioinformatics Group
• Lance Davidow, Jonathan Urbach