Transcript introduction_to_micr..

Introduction to Gene
Expression Analysis
Phillip Lord
Resources
http://www.ebi.ac.uk/microarray/biology_intro.html
http://www.mged.org/
http://www.tigr.org/tdb/microarray/
Microarray Bioinformatics
Dov Stekel
Product Details:
Paperback 280 pages (September 8, 2003)
Publisher: Cambridge University Press
Language: English
ISBN: 052152587X
How do we measure gene
expression?
• Oldest technique is to
look at a phenotype.
• In this case, the ura4+
gene from S.pombe
• Most other techniques
based on hybridisation.
– Northern Blot
– Quantative RT-PCR
Microarray analysis
• Whole genome sequencing makes it possible to
predict the entire gene complement
• Various technologies have built on this
knowledge to produce systems that will monitor
the expression (usually transcription) at the
whole genome level
– Measurement of global transcription is called
transcriptomics
• Come by a variety of names – gene chips,
arrays, DNA arrays. Can be somewhat
confusing what is actually being described.
• Not to be confused with Genotyping Microarrays
Generating Microarrays
• There are many different systems for
generating microarrays
– spotting
• original technology, now rather old
• good for “one off” arrays
– in-situ synthesis
• newer, more reproducible
• expensive first time around, then cheaper
Spotting
• Synthesize DNA, spot onto glass slides,
fix.
A Spotting Robot
The head
A Spotting Pin
taken from Stekel, 2003
In-situ synthesis
•
•
•
•
Uses chemically protected nucleotides
Specific spots are “de-protected”
Can then extend these oligos
Different techniques for deprotection
Masked Synthesis
•
Uses masks much like silicon
chip production
•
Masks are expensive
•
Good for bulk production,
standard arrays
Photo deprotection
• A light source is used to deprotect
oligos
• Essentially, this is much the same
as an LCD projector.
from Stekel, 2003
InkJet Synthesis
• An InkJet head is used to place
nucleotides at the appropriate place
on the array
An experiment
RT with Cy3 dCTP
Two Samples
Combine into
single sample
RT with Cy5 dCTP
Hybridise to Microarray
Hybridisation
from Stekel, 2003
Detection
• Finally, the hybridisation extract must be
detected
• The technology is related to desktop
scanners, but more sensitive.
• Usually produces a TIFF file
from Stekel, 2003
The end result
from Stekel, 2003
Problems
• We are looking for variability between the
expression of different genes.
• There are many (many!) other sources of
variability
• Most microarray analysis is about trying to
normalise these sources of variability,
leaving biological variability
Artifacts
The Jolly Green Giant
The Yellow Splodge Peril
Space Invaders
Solutions
• Removing Sections
• Background Subtraction
• Start Again
Feature Recognition
• Not all spots are equal – different sizes,
different shapes.
• Identifying the exact scope of the spot on
an array can therefore be hard.
• Often solved in the initial detection of
spots.
Spot Detection
The Doughnut
A general disaster
The basic solution to this is to not use
circular spots for detection. There are
a variety of edge detection algorithms,
or manual tools which work.
An experiment
RT with Cy3 dCTP
Two Samples
Combine into
single sample
RT with Cy5 dCTP
Hybridise to Microarray
Channel Variability
• Cy3/Cy5 dyes have different properties.
• So do the lasers at different frequencies.
• So do the photomultipliers which detect
them.
Within Slide Variability
• Slides often have imperfections, either from
spots, or background
• Gaps are not uncommon, neither are
chromatic effects
Inslide Normalisation
Between slide variability
• Results between different slides are not
directly comparable.
• Results between different experiments are
not directly comparable.
Further work
Smith JR, Choi D, Chipps TJ et al. Unique gene expression profiles of
donor matched human retinal and choroidal vascular endothelial cells.
Invest Ophthalmol Vis Sci 2007;48:2676-2684.
Chi JT, Chang HY, Haraldsen G et al. Endothelial cell diversity
revealed by global expression profiling. Pro Nat Acad Sci
2003;100:10623-10628.