Transcript MicroArray -- Data Analysis

MicroArray -- Data Analysis
Bioinformatics - 10p, October 2001
Cecilia Hansen
&
Dirk Repsilber
1
Overview
• Bio-Informatic motivation
• Monday
–
–
–
–
MA experimental basic (C)
MA data analysis (D)
Introduction to lab 1 (C)
lab 1 (C&D)
• Tuesday
– Introduction to lab 2 (D)
– lab 2 (D&C)
2
Bio-Informatic motivation
• Functional Genomics
– adaptation and development
– gene regulation & gene function
– ”Transcriptome” (diagnostical purposes)
• Analysis of large-scale noisy data
– data management
– statistical tools
– associating further biological knowledge (database-links)
3
Overview
• Bio-Informatic motivation
• Monday
–
–
–
–
MA experimental basic
MA data analysis
Introduction to lab 1
lab 1
• Tuesday
– Introduction to lab 2
– lab 2
4
experimental basics
Target (solid phase)
1) Genomic DNA or cDNA clones with known sequence.
2) PCR products are spotted on a glas slide in an organised way.
3) With a good printing device you can print at leased 10 000
spots (genes) on one microscope glas slide.
4) This gives you a matrix of spots, where every spot represents
only ONE gene.
***** ***** *****
***** ***** *****
***** ***** *****
*****
*****
*****
*****
*****
*****
5
experimental basics
Probe
control
RNA is isolated from the cells
t=0
• mRNA is a very instable
molecule!
t=0
t=1
sample
t=2
t=3
...
t=7
• The amount of mRNA is in
direct relation to the expression
level of the gene.
Make cDNA from mRNA with reverse transcriptase and
incorporate fluorophores Cy3 and/or Cy5.
6
experimental basics
Hybridization
1) Mix equal amounts ( 1:1 ) of Cy3 and Cy5 samples and
hybridize it on to the array.
green > red
green = red
green < red
2) The probe will bind to the complementary target to the
array.
3) A laser scanner is then used to make the fluorophores
to emit light with different wavelength. This will give
us a picture of the array with spots in green, yellow and
red.
7
MA data analysis
• How do MA data look like ?
• What are the specific questions ?
• Methods
8
Overview
• Bio-Informatic motivation
• Monday
–
–
–
–
MA experimental basic
MA data analysis
Introduction to lab 1
lab 1
• Tuesday
– Introduction to lab 2
– lab 2
9
MA data analysis
an expression profile like this for each gene :
[mRNA] ~ Cy5/Cy3 = r
5_
1
_
up-regulation
induction
down-regulation
repression
0
Start of experiment
time / h
10
MA data analysis
Co-Regulation -- Inference
function: expression:
Differential
• Which genes are differentially of
Genes
belonging
expressed ?
Comparing
the to the same
pathway
are
often
Transcriptomes for two
• Which genes are expressed in a showing the same regulatory
different
biological
”similar” way when comparing
Expression-Fingerprinting:
patterns (profiles) for a variety
to expression profiles of genes
samples
(e.g.
control,
Often
in medical
applications
of biological
situations
(or in a
with known function?
heat-shock)
are
it
is of
interestyou
to characterize
time
series).
(co-regulation)
Reverse Engineering:
the
biological
status
of cells,
interested
in
the
subset
of
Hence,
as
a
hypothesis,
genes
Using expression data to
e.g.
thewhich
severeness
of showing
tumor
of
unknown
function
• patterns of expression
genes
are
infer regulatory interactions
cells,
toregulatory
be able to behaviour
respond
(diagnostic ”Fingerprinting”)
similar
expressed
on
different
between a number of genes
with
the genes
right therapy.
as
some
of
”known”
responsible
for
a
certain
levels
• Reverse Engineering of
function
may
have or
a similar
adaptation
process
genetic networks
(up-/
down-regulated).11
function.
developmental process.
MA data analysis - methods
12
MA data analysis - methods
GeneSpring
R - sma_package
Mean-Mean-Plot
40000
Excel
35000
Mean(Ch2)
30000
25000
20000
15000
10000
5000
0
0
10000
20000
30000
40000
Mean(Ch1)
13
MA data analysis
Biological question
Differentially expressed genes
Classification etc.
Experimental design
Microarray experiment
lab 1
Image analysis
lab 2
Normalization
Description
Testing
Clustering
Biological verification
and interpretation
Discrimination
14