DNA microarray technique enables one to analyze the expression of many genes in a single reaction quickly and in an efficient manner. This technique has been elaborately described in this presentation
2. INTRODUCTION
● DNA microarray is a uniquely efficient method for simultaneously assessing the
expression levels of thousands of genes.
● DNA microarray consists of a solid surface, known as DNA chip, onto which
DNA molecules have been chemically bonded.
● The purpose of a microarray is to detect the presence and abundance of labelled
nucleic acids in a biological sample which will hybridise to the DNA on the
array via Watson–Crick duplex formation, and which can be detected via the
label.
.
3. CONTENTS
❏ Microarray probe preparation
❏ Using microarray
❏ Image processing
❏ Normalisation of the data
❏ Quantification of variables
5. 1. Robotic spotting
In spotted microarrays, the probes are oligonucleotides, cDNA or small fragments of
PCR products that correspond to mRNAs.
The probes are synthesized prior to deposition on the array surface and are then
spotted onto glass.
A common approach utilizes an array of fine pins or needles controlled by a robotic
arm that is dipped into wells containing DNA probes and then depositing each probe
at designated locations on the array surface.
6. Oligonucleotide Probe Design
Select 3’ portion of the
gene , mask the repeat
sequences and generate
all possible oligos
Check melting
temperature of the
probes
Select gene for
oligonucleotide synthesis
Check sequence
homologies and remove
bad probes
Check secondary
structure of the probe
Select the appropriate
probe for microarray
7. There are two main types of spotted array which can be subdivided in two ways:
❖ Type of DNA probe: The DNA probes used on a spotted array can either be
polymerase chain reaction (PCR) products or oligonucleotides.
❖ Attachment chemistry of the probe to the glass: Via covalent or non covalent
bond.
8. ➢ With covalent attachment, a primary aliphatic amine (NH2) group is added to
the DNA probe and the probe is attached to the glass by making a covalent
bond between this group and chemical linkers on the glass.
➢ The amine group can be added to either end of the oligonucleotide during
synthesis, although it is more usual to add it to the 5’ end of the oligonucleotide.
10. 2. In-situ synthesis of oligonucleotide arrays
These arrays are fundamentally different from spotted arrays:
● Instead of pre-synthesising oligonucleotides, oligos are built up base-by-base on
the surface of the chip. This takes place by covalent reaction between the 5’
hydroxyl group of the sugar of the previous nucleotide attached and the
phosphate group of the next nucleotide.
● Each nucleotide added to the oligonucleotide on the glass has a protective group
on its 5’ position to prevent the addition of more than one base during each
round of synthesis. The protective group is then converted to a hydroxyl group
either with acid or with light before the next round of synthesis.
11. Methods For Deprotection
The three main technologies for making in-situ synthesized arrays:
A. Photodeprotection using masks: This is the basis of the Affymetrix®
technology.
B. Photodeprotection without masks: This is the method used by Nimblegen and
Febit.
C. Chemical deprotection via inkjet technology: This is the method used by
Rosetta, Agilent and Oxford Gene Technology.
12. A.Affymetrix technology
● Affymetrix arrays use light to convert the
protective group on the terminal nucleotide
into a hydroxyl group to which further bases
can be added.
● The light is directed to appropriate features
using masks that allow light to pass to some
areas of the array but not to others.
14. B. Maskless Photodeprotection Technology
● This consists of a large number of
mirrors embedded on a silicon chip,
each of which can move between two
positions: one position to reflect light,
and the other to block light.
● At each step, the mirrors direct light to
the appropriate parts of the array.
● an array of mirrors is computer
controlled and can be used to direct light
to appropriate parts of the glass slide at
each step of oligonucleotide synthesis.
15. Image Courtesy: Melissa B. Miller and Yi-Wei Tang; ‘Basic Concepts of Microarrays and Potential
Applications in Clinical Microbiology’ ; Oct. 2009, p. 611–633; doi:10.1128/CMR.00019-09.
16. C. Inkjet Array Synthesis
● This technology uses chemical
deprotection to synthesize the
oligonucleotides.
● The bases are fired on to the array
using modified inkjet nozzles,
which, instead of firing different
colored ink, fire different
nucleotides.
● At each step of synthesis, droplets
of the appropriate base are fired
onto the desired spot on the glass
slide.
Image Courtesy: Melissa B. Miller and Yi-Wei Tang; ‘Basic Concepts of Microarrays and Potential
Applications in Clinical Microbiology’ ; Oct. 2009, p. 611–633; doi:10.1128/CMR.00019-09.
17. Image Courtesy: Melissa B. Miller and Yi-Wei Tang; ‘Basic Concepts of Microarrays and Potential
Applications in Clinical Microbiology’ ; Oct. 2009, p. 611–633; doi:10.1128/CMR.00019-09.
FIG: inkjet technology
18. Spot Quality
Inkjet arrays tend to be of the
highest quality, with regular,
even spots.
Spotted arrays produce
spots of variable size and
quality.
Affymetrix arrays the features
are rectangular regions
19. USING MICROARRAY
There are four laboratory steps in using a microarray to measure gene expression in
a sample.
I. Sample preparation and labeling
II. Hybridization
III.Washing
IV.Image acquisition
21. I. Sample preparation and labelling
● The first step is to extract the RNA from the tissue of interest.
● With most technologies, it is common to prepare two samples and label them
with two different dyes, usually Cy3 (excited by a green laser) and Cy5 (excited
by a red laser).
● The samples are hybridized to the array simultaneously and incubated for
between 12 and 24 hours at between 45 and 65˚C.
● The array is then washed to remove sample that is not hybridized to the
features.
22. II. hybridization
● Hybridization is the step in which the DNA probes on the glass and the labeled
DNA(or RNA) target form hetero duplexes via Watson–Crick base-pairing.
● It is affected by many conditions, including temperature, humidity, salt
concentrations, formamide concentration, volume of target solution and
operator.
23. III. washing
● After hybridization, the slides are washed, this ensures that the only labeled
target on the array is the target that has specifically bound to the features on
the array and thus represents the DNA that we are trying to measure.
● It also increase the stringency of the experiment by reducing cross-
hybridization.
24. IV. Image Acquisition
● The slide is placed in a SCANNER, which is a
device that reads the surface of the slide.
● Each pixel on the digital image represents the
intensity of fluorescence induced by focusing the
laser at that point on the array.
● The dye at that point will be excited by the laser
and will fluoresce; this fluorescence is detected
by a photomultiplier tube (PMT) in the scanner.
FIG. Working of scanner
25. IMAGE PROCESSING
● The image of the microarray generated by the scanner is the raw data of
experiment.
● Computer algorithms, known as feature extraction software, convert the image
into the numerical information that quantifies gene expression.
26. Feature extraction
● The first step in the computational analysis of microarray data is to convert the
digital TIFF(tagged image file format) images generated by the scanner into
numerical measures of the hybridization intensity of each channel on each
feature. This process is known as feature extraction.
27. Steps for image processing
There are four steps:
1. Identify the positions of the features on the microarray.
2. For each feature, identify the pixels on the image that are part of the feature.
3. For each feature, identify nearby pixels that will be used for background
calculation.
4. Calculate numerical information for the intensity of the feature, the intensity of
the background and quality control information.
28. 1. Identifying the Positions of the Features
The features on most microarrays are arranged in a rectangular pattern. However,
the pattern is not completely regular.
The features on the array are arranged in grids, with larger spaces between the
grids than between the features within each Grid.
30. 2. Identifying the Pixels That Comprise the
Features
The next step in the feature extraction procedure is called segmentation This is the
process by which the software determines which pixels in the area of a feature are
part of the feature So their intensity will count towards a quantitative measurement
of intensity at that feature.
31. There are four commonly used methods for segmentation:
a. Fixed circle
b. Variable circle
c. Histogram
d. Adaptive shape
Different software packages implement different segmentation algorithms and some
packages implement more than one algorithm, which gives the user the option to
compare different algorithms on the same image.
33. 3. Background calculation
a. ScanAlyze: the region is adjacent to the feature. This will be inaccurate if the feature is larger than
the fixed size of the circle used for segmentation.
b. ImaGene: there is a space between the feature and the background. This is a better method than
a.
c. Spot and GenePix: the background region is in between the features. This is also a good method.
34. 4. Calculation of Numerical information
After determining the pixels representing each feature, the image-processing
software must calculate the intensity for each feature.
Image-processing software will typically provide a number of measures:
➢ Signal mean: The mean of the pixels comprising the feature.
➢ Background mean: The mean of the pixels comprising the background around
the feature.
➢ Signal median: The median of the pixels comprising the feature.
➢ Background median: The median of the pixels comprising the background.
35. ➢ Signal standard deviation: The standard deviation of the pixels comprising the
feature.
➢ Background standard deviation: The standard deviation of the pixels comprising
the background.
➢ Diameter: The number of pixels across the width of the feature.
➢ Number of pixels: The number of pixels comprising the feature.
➢ Flag: A variable that is 0 if the feature is good, and will take different values if
the feature is not good.
36. DATA NORMALISATION
Normalization is a general term for a collection of methods that are directed at
resolving the systematic errors and bias introduced by the microarray experimental
platform
37. ❖ Data Cleaning and Transformation: Deals with cleaning and transforming the
data generated by the feature extraction software before any further analysis
can take place.
❖ Within-Array Normalization: Allow for the comparison of the Cy3 and Cy5
channels of a two-colour microarray. This section is only relevant for two-
colour arrays.
❖ Between-Array Normalization: Describes methods that allow for the
comparison of measurements on different arrays. This section is applicable both
to two-colour and single channel arrays, including Affymetrix arrays.
38. FIG: weighted smoothed medians of difference of expression values for the human brain tissue data. A) without
normalization, B) after median normalization, C) after quantile normalization and D) after cyclic loess.
39. QUANTIFICATION OF VARIABLES
FIG: Sources of variability in microarray experiment
● Experimental variability is measured
with calibration experiments
● Population variability is measured with
pilot studies
40. Steps for measuring variability
➔ For each set of replicates (features , hybridizations or individuals), calculate the
mean of the replicates.
➔ For each replicate, calculate the deviation from the mean by computing the
difference between the intensity of the replicate and the mean of the set of
replicates.
➔ Produce MA plots of the deviations against the mean to check that the
variability is independent of intensity.
41. ➔ An appropriate linear or non-linear normalization can be applied to the
deviations.
➔ Calculate the standard deviation of the error distribution using all of the
replicates.
➔ If the variability depends on the intensity, then partition the data into different
intensity ranges and calculate a standard deviation for each partition.
42. ➔ If the log-normal assumption is true, then the deviates should be distributed as
a normal distribution. This can be checked by plotting a histogram of the
deviates.
➔ Convert the standard deviation to a percentage coefficient of variability by
multiplying by ln(2) and applying Equation.
43. REFERENCES
1. Dari Shalon, Stephen J. Smith, and Patrick O. Brown, A DNA Microarray System for Analyzing
Complex DNA Samples Using Two-color Fluorescent Probe Hybridization, Genome Res. 6: 639-645,
2017.
2. John Quackenbush,Microarray data normalization and transformation, nature genetics supplement,
Vol 32, 2002.
3. Melissa B. Miller and Yi-Wei Tang Basic Concepts of Microarrays and Potential Applications in
Clinical Microbiology, Oct. 2009, p. 611–633; doi:10.1128/CMR.00019-09.
4. Richard P. Auburn, David P. Kreil, Lisa A. Meadows, Bettina Fischer, Santiago Sevillano Matilla and
Steven Russell, Robotic spotting of cDNA and oligonucleotide microarrays, TRENDS in Biotechnology
Vol.23 No.7 July 2005.
5. Youlan Rao, Yoonkyung Lee, David Jarjoura, Amy S. Ruppert, Chang-gong Liu, Jason C. Hsu, and
John P. Hagan, A Comparison of Normalization Techniques for MicroRNA Microarray Data,
Statistical Applications in Genetics and Molecular Biology, Vol. 7, 2008.