High-throughput genotyping

1
Microarrays overview, one and two colour systems
Richard Talbot Roslin Institute and
R(D)SVS University of Edinburgh
2
Microarray uses

• Definition
• A microarray is a high density of DNA probes of
  known sequence arrayed across a solid surface
• Uses
• Gene Expression
• Comparative Genome Hybridisation (CGH)
• Single Nucleotide polymorphisms (SNP)
• DNA methylation
• Copy Number variation (CNV)

3
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

4
Experimental model

• Gene expression in the developing limb
• model During development different genes are
  expressed in a choreographed order for the
  spatial patterning of tissues. Chick limb buds at
  stage 20 (3 days) show patterning of Sonic
  Hedgehog a major transcription factor controlling
  the anterior-posterior axis in the limb bud.

5
Experimental questions

1. What are the genes associated with patterning in
   the limb bud?
2. What are the genes influenced by the presence of
   Sonic Hedgehog protein?

6
Experimental plans
Question 1
Posterior expression
Anterior expression
Vs
Question 2
Anterior bead expression
Posterior expression
Vs
Anterior expression
Anterior Bead expression
Vs
Anterior Gli3 expression
Posterior expression
Vs
Anterior expression
Anterior GLi3 expression
Vs
7
Reference design
One Colour Experiment
Reference
Anterior bead expression
Anterior expression
Posterior expression
Anterior Gli3 expression
8
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

9
Types of Microarrays

• Probe type
• cDNA
• Oligonucleotide short (lt35 bases)
• Oligonucleotide long (gt35 bases)

10
Oligonucleotide synthesis
11
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

12
RNA quality
13
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

14
RNA Labelling
Indirect
Direct
Total RNA
Cy3-CTP or Cy5-CTP
Aminoallyl-UTP
Reverse transcription
cDNA
cDNA
Dye Coupling Cy3 or Cy5
15
RNA amplification
Total RNA
cDNA
Reverse transcription
Remove RNA and ds-cDNA
cRNA
In Vitro transcription
16
Affymetrix Single Channel Microarray Experiment
17
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

18
Microarray Experiment
Control RNA
Treatment RNA
Reverse Transcription
Dye Coupling
Cy5
Cy3
Emission
Excitation
Mix on Slide
Scanning
Hybridise
19
Hybridisation

• Temperature control
• Mixing
• Multi-chamber
• Loading sample
• Washing

20
Scanning

• Set laser power
• Set PMT gain or auto adjust
• Set scan area
• Apply griding file
• Align griding file
• Flag problem features
• Extract Values

21
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

22
Sources of variation 1

• Microarray production
• Print-pin anomalies
• Variations in printed probe quantities even with
  the same pin
• Chip batch variation (due to many sources of
  unknown variations)
• Differences in sequence length of the immobilized
  DNA
• Variations in chemical probe attachment levels to
  the slide

• mRNA
• Differences in conditions.
• Differences between experimental subjects within
  the same covariate level.
• Differences between samples from the same
  subject.
• Variation in mRNA extraction methods from
  original sample.
• Variations in reverse transcription.
• Differences in PCR amplification.
• Different labelling efficiencies.

23
Sources of variation 2

• Scanning/imaging
• Different scanners.
• Different photo-multiplier gain.
• Different spot-finding software.
• Different grid alignments.
• Different laser power

• Hybridization process
• Different dye sensitivities.
• Inequalities in the application of mRNA to the
  slide.
• Variations in the washing efficiencies of
  non-hybridized mRNA on the slide.
• Other differences in hybridization parameters,
  such as
• temperature
• experimenter
• time of the day

24
Normalisation (two dye)

• Used to remove systematic bias form the image
  data
• Methods
• Global
• All genes
• Housekeeping/spike genes
• Loess
• Global
• Pin tip
• Spatial

25
Normalisation (one dye)

• Method depends on slide system
• Affymetrix
• MAS5.0
• RMA
• Plier
• Agilent
• Spatial normalisation

26
Differential expression

• T-Test
• Multiple testing - False Discovery Rate (FDR)
• ANOVA

27
Microarray experiment

• Define the questions
• Choose microarray
• RNA preparation
• Target labelling
• Array Hybridisation
• Image scanning and analysis
• Statistical analysis
• Reporting results

28
Minimal Information About a Microarray Experiment

• Database of microarray experiment

Protocols
Experiment
Samples
ARRAY
Hybridisation
Data Analysis
29
Gene Expression Data Repositories

• Major public repositories
• ArrayExpress (EBI) - http//www.ebi.ac.uk/microarr
  ay-as/ae/
• GEO (NCBI) - http//www.ncbi.nlm.nih.gov/geo/
• CIBEX (DDBJ) http//cibex.nig.ac.jp/index.jsp
• Submission tools
• Tab2Mage - http//www.ebi.ac.uk/cgi-bin/microarray
  /tab2mage.cgi
• GEOarchive - http//www.ncbi.nlm.nih.gov/projects/
  geo/info/spreadsheet.htmlGEOarchive
• Cibex - http//cibex.nig.ac.jp/submission/index.js
  p

30
After the Gene list

• Electronic
• Annotation
• Clustering
• Pathways
• Text mining
• Networks

• Practical
• qPCR
• In Situ hybridisation
• Northern blot
• Western Blot
• Immunohistochemistry
• siRNA
• Pharmacology

31
Conclusions

• Several successful microarray platforms are
  available
• Choice of platform is based on experimental
  design and cost
• It is key to know the questions being asked
• Many steps in a microarray experiment contribute
  to the overall variance
• Normalisation can be used to remove technical
  variance
• Microarrays are a tool to answer a specific
  question have a plan on how to confirm the answer.