This presentation will take you through the concept of microarray and its application in detail.

1. Purnima Kartha. N
1

2. • A microarray consists of a solid surface to which biological molecules are
arranged in a regular pattern.
• Applicable in the fields of DNA, proteins, peptides and small molecules like
metabolites and drugs.
INTRODUCTION
2

3. • Orderly arrangement of thousands of
identified sequenced genes printed on an
impermeable solid support, usually glass, silicon
chips or nylon membrane.
• Thousands of spots each representing a single
gene and collectively the entire genome of an
organism.
• Measurement of Gene Expression.
DNA MICROARRAY
3

4. PRINCIPLE
 Hybridization between two DNA strands
 Microarrays use relative quantitation : Intensity of a feature is compared to
the intensity of the same feature under a different condition, and the identity of
the feature is known by its position.
4

5. 1. Glass DNA microarrays which involves the
micro spotting of pre-fabricated cDNA fragments
on a glass slide.
2. High-density oligonucleotide microarrays
often referred to as a "chip" which involves in situ
oligonucleotide synthesis.
 Affymetrix : by photolithography
 Agilent: Inkjet printing technology
TYPES OF MICROARRAYS
5

6.  The experimental steps involved include:
STEPS
6

7. 1. ARRAY PRINTING
• The microarray slide is uniformly coated with a chemical compound that will
interact with and immobilize nucleic acids, irreversibly binding them to the
surface.
• Nucleic acids are deposited on the slide by contact printing, and treated with
UV light or baking at 80°C to crosslink them to the slide surface.
• The printed slides are stored desiccated at room temperature, in the dark,
until required for experimentation.
7

8. PROBES
8
 The selection of probe sequences for a
microarray design depends upon the use
envisaged for the array.
 Specific probes can be designed against genes,
transcripts or portions of transcripts.
1. Clone sets
2. cDNA library preparation
3. Specific DNAs amplified.
 All clones purified by gel filtration/ precipitation
to reduce unwanted salts.
 Selection of probes requires balancing of 4
criteria during their production: Sensitivity,
Specificity, Noise, and Bias.

9. Printing PCR products onto glass slides:
• Printing involves the sequential transfer of individual PCR products from the
plates to defined areas of glass slides.
• Glass slides pre-coated with poly lysine, amino silanes or amino-reactive
silanes - to increase the hydrophobicity of the slide, improving the adherence
of the deposited DNA and minimising spreading.
• Few nLs of each DNA is deposited onto each slide, resulting in formation of
spots of 50-150µM diameter.
9

10. 96-well plate
Contains cDNA
probes
Glass slide
Array of bound cDNA probes
cDNA clones
Print-tip
Print-tips collect
cDNA from wells
10
Printing of cDNA microarrays

11. 11
Photodeprotection using masks:
Affymetrix (Photolithography)
Photodeprotection without masks:
Nimblegen, Febit
Inkjet Technology: Agilent, Oxford
Gene Technology.
Printing of High-density oligonucleotide microarrays

12. Post Processing of slides:
• DNA is usually cross-linked to the glass slides by treating with UV light or
baking at 80°C, and residual amines are blocked be reaction with succinic
anhydride.
• As a final step, a proportion of the deposited DNA is rendered into single
stranded form available for hybridisation by heat denaturation.
• The printed slides are stored desiccated at room temperature, in the dark,
until required for experimentation.
12

13. 2. SAMPLE PREPARATION
In sample preparation, RNA from the host organism is isolated, converted
to cDNA and labelled with dyes before hybridization to the array.
1. RNA extraction from the tissue of interest.
 Quantity, and integrity of the total or mRNA used is important
 Numerous methods for RNA isolation available:
 Use of Trizol®, and other phenol-based methods
 Commercially available RNA isolation kits (eg: Ambion, Qiagen, and Promega.)
provide rapid and reliable RNA extraction.
2. cDNA production: convert the RNA into a labelled form for hybridization.
 This most typically involves a reverse transcription step.
13

14. Labelling
 Combination of Cy3 (excited by green laser) and
Cy5 (excited by red laser) has been used most
frequently.
 Relatively stable in light
 Incorporated efficiently into cDNA
 Wide separation in excitation and emission spectra
 2 samples are hybridised to the arrays, one labelled
with each dye, allows simultaneous measurement of
both samples.
 2 widely used methods of labelling cDNA:
 Direct & Indirect labelling
14

15. Direct labelling:
 A Dye conjugated ntd incorporated directly into cDNA by RT enzyme.
 By using dNTPs that have dye molecule directly coupled to the base, with
cyanine 3-dCTP (Cy3-dCTP) and cyanine 5-dCTP (Cy5-dCTP).
 Conjugates of the Alexa dyes, Alexa555 and Alexa647, the spectral
analogues, fluorescent and photo-stable and are therefore the most
commonly used alternatives.
.
Ad: Quick and simple, requiring relatively few steps, and therefore are easy to scale
up for high throughput.
Disad: Require high amount of RNA (approx 25–100 mg total RNA) for
labelling reaction. The bulky dye-coupled nucleotides reduce the efficiency of the
reverse transcriptase and lead to dye bias. 15

16. 16

17. Indirect labelling:
 The RT enzyme incorporates an amino-allyl dNTP into the cDNA instead of
Cy-dCTP.
1. Aminoallyl-dNTP is added to ntd mix in the RT reaction to produce first strand
cDNA
2. After first strand synthesis, an amine-reactive Cy dye is chemically coupled to the
aminoallyl groups, thus labelling the cDNA.
The CyDyes have NHS (N-Hydroxysuccinimide) esters that react with the aminoallyl
groups of the cDNA.
 Ad: Less steric hindrance by smaller group (an aminoallyl-dNTP).
17

18. 18

19. Other Labelling Methods:
The sample RNAs hybridized to the complementary probes on the array are
detected by incubating the array in a colloidal gold solution.
 The +vely charged gold particles are attracted to -vely charged phosphate
groups in the backbone of the target, resulting in precipitation of nano-gold
particles.
 The amount of precipitation proportional to the amount of bound target
RNA.
 Ad: Instead of an expensive confocal scanner, a relatively inexpensive flatbed
scanner can used to detect the gold precipitate.
19

20. 3. MICROARRAY HYBRIDISATION
During the hybridization reaction, labelled targets interact with the tethered probes
due to sequence complementarity.
o Appropriate hybridization conditions are critical to ensure correct measurement.
o The hybridization procedure involves several steps:
• The arrays are blocked to minimize background.
• The labelled target is added to the array at a specific temperature to allow
complementary sequences to anneal.
• The arrays are washed to remove unbound or weakly hybridizing material.
20

21. Blocking
 Before hybridization, the array is treated to prevent nonspecific interactions
between the nucleic acid in the labelled sample and the array surface.
 Different blocking methods have been described by the chemistry of slide
coating.
 After blocking, double stranded DNA arrays are boiled to denature the DNA
and thus enhance their availability for hybridization.
polyL-lysine arrays need to have exposed amines blocked to prevent binding of
labelled material.
Achieved with a mixture of succinic anhydride, 1,2-methyl pyrrolidinone and
sodium borate. Succinic anhydride reacts with and caps the amines before the
excess DNA from the printed probes leaches from the spot area and binds nearby
exposed lysines.
21

22. Hybridization
Hybridization depends on the ability of the labelled target to anneal to a
complementary probe strand tethered to the array.
 This occurs just below Tm of the target–probe duplex.
 The main factors affecting are temperature, pH, monovalent cation concentration
and the presence of organic solvents.
 Hybridization solution: Contain a high concentration of salts, detergents,
accelerants, and buffering agents. The most common components of solutions
are:
 Sodium chloride and sodium citrate (SSC)
 Formamide and dithiothreitol (DTT)
 Dextran sulfate
 EDTA
 Sonicated salmon sperm DNA, polyA, Denhardt’s solution
22

23. After-Hybridisation Washing:
 To remove unbound target and any target loosely bound to imperfectly
matched sequences.
 For good quality arrays, it is essential that both hybridization and washing is
uniform across the array and that the surface is evenly dried before scanning.
LABEL
3XSSC
HYB CHAMBER
ARRAY
SLIDE
LIFTER SLIP
SLIDE LABEL
23

24. 4. DATA ACQUISITION &ANALYSIS
• Gene expression levels are evaluated by measuring the amount of reference
and test probe that binds to each arrayed cDNA.
• Fluorescence is detected on arrays by means of a scanner or reader and
saved as a digital image.
• These data are then imported into software that converts the fluorescence into
pixels that can be counted.
• Following background subtraction and normalisation, expression ratios can be
calculated.
24

25. 1. Scanning
o A microarray scanner uses a light source to excite the fluorophores present on
the sample molecules and then detects the emitted light, which is most typically
stored as a 16-bit tiff (tagged image format) files for each wavelength scanned.
o Each image is composed of a matrix of pixels where each pixel represents the
fluorescence intensity of a small area of the array.
o This digital image represents the ‘raw data’ from which the fluorescent signal of
each array element will subsequently be quantified and the expression level of
each target inferred.
25

26. Detector
PMT
Image
Cy5: 635nm
Cy3: 532nm
26

27.  Scanners, according to the type of excitation and detection technology they utilize.
Scanners
Laser
Scanners
CCD based
Scanners
Laser Scanners:
• Use narrowband laser illumination to excite
fluorophores and capture the resulting
fluorescence with a PMT detector.
• Any slight movement in the position of the array
or difference in the starting point of the scanning
head can lead to mis-registration.
27

28. CCD (charge-coupled device) Scanners:
 The excitation source for a CCD detector is broad-spectrum white light,
(xenon or mercury lamp), which is filtered to select the excitation wavelength.
 The filtered light illuminates a large area of the array and the fluorescent
emission from the entire field of view is collected by a stationary CCD.
 A CCD chip is an array of semiconductor devices, or camera pixels, where
each camera pixel stores an electrical charge generated by light from the
emission filter.
 This charge is proportional to the intensity of the light, or number of
photons, that reaches the semiconductor.
 Electronic circuitry on the camera converts pixel electron counts from the
CCD chip into a digital signal that represents the intensity of each pixel.
28

29. • CCD scanners tend to capture multiple
images from different areas of the array
and then stitch them together to create a
single image.
• Disadvantage: Any imprecision in the
stitching, or photo bleaching due to
repeated exposure of overlapping
regions, will result in inaccurate and
uneven signal quantification.
• Advantage: Data from all wavelengths
simultaneously collected and therefore
no mis-registration issues. 29

30. 2. Image Data Generation:
• The images obtained from the scanner are imported onto software that
converts the 2 scans into pseudo-coloured images that can merged.
• Software for this purpose is available from a no. of sources, both academic and
commercial.
• Gridding
• Background sampleing
• Target intensity extraction
• Signal filtering
• Ratio Calculation
• Processing expression data
30

31. 2.1. Gridding:
 Define the location of the different arrayed cDNA spots
by overlaying the image with a grid in which each spot is
contained within a defined circle.
 The process of gridding is aided by the highly defined
arrangement of spots produced by the robotic printing of
arrays.
31
2.2. Background sampling:
 The regions where probe is located have different brightness characteristics
in comparison to non-spot regions without probe.
 Separating the foreground spot signal from the background.
 To calculate background, median intensities of the background pixels can be
determined separately for the corresponding Cy3 and Cy5 image.

32. 32

33. 2.3. Target intensity extraction:
 To determine spot intensity, the mean intensity of the pixels contained within
the spot circle are determined for corresponding Cy5 and Cy3 image.
 Subtraction of background pixels for the corresponding element then yields a
net intensity value for the particular arrayed DNA.
33
2.4. Signal filtering:
 Pixel intensities for unexpressed or poorly expressed genes would be
expected to be close to the background intensity of Cy3 and Cy5 image.
 To distinguish such genes, data may be filtered so that only those genes
expressed above a certain level are analysed further.
 Typically spot intensities that are less than 2 fold above background can be
excluded.

34. 2.5. Ratio Calculation:
 Calculate the ration of the background corrected intensity for the Cy5 image to
that for the Cy3 image for each spot.
 Accuracy of this measurement can vary, in particular on arrays with high
background.
M = log2 R/G = log2R - log2G
M < 0, gene is over-expressed in green-labeled sample
compared to red-labeled sample.
M = 0, gene is equally expressed in both samples.
M > 0, gene is over-expressed in red-labeled sample compared
to green-labeled sample.
34

35. 2.6. Processing expression data:
 The ratios are normalised to exclude any bias towards one of the two
probes.
 Normalization is carried out based on the assumption that only a small
proportion of genes will be differentially expressed among the thousands of
genes present in the array and/or that there is symmetry in the up- and
down-regulation of genes.
 Normalised data can then be ranked to identify expression changes between
the reference and test condition.
 Finally data can be organised to identify similar expression patterns.
35

36. 36

37. APPLICATIONS OF DNA
MICROARRAY TECHNOLOGY
37

38. GENE EXPRESSION MONITORING
 Measurement of absolute levels of expression for each represented gene by
quantifying the amount of targets that hybridise with the arrayed probes.
 Expression arrays can be used to catalogue which genes are expressed in a
particular cell or tissue sample. Allow identification of a molecular fingerprint
to characterise cells in different stages of differentiation.
 Expression arrays can also be used to study dynamic changes in gene
expression over time.
 DNA microarray technology offers the possibility of high-throughput
systematic analysis of the transcriptome in one experiment. 38

39. DIAGNOSTIC ARRAYS
39
 Diagnostic assays are playing an increasingly
important role in the treatment of diseases.
 A rapid, accurate, and reliable diagnostic
method allows identification of the disease
for suitable therapy, which consequently can
reduce the mortality rate and also the cost of
treatment.

40. MICROBIAL DETECTION AND
IDENTIFICATION
 The most commonly used gene targets are16S bacterial and 28S fungal and
intergenic transcribed spacers (ITSs) in rRNA genes, and microarray technology
has been incorporated to compensate for the time-consuming sequencing
identification procedure.
40
Microarray targeting the 16S rRNA gene developed for the detection of a panel
of 40 predominant human intestinal bacterial pathogens in human fecal samples.
Rapid diagnosis of bloodstream infections caused by common bacterial
pathogens in the paediatric and general populations.
Use of microarrays to identify pathogenic yeasts and molds by targeting the ITS
regions in fungal rRNA genes

41.  Allows simultaneous identification of candidate biomarkers by analysing
differentially expressed genes under comparative conditions such as healthy vs.
diseased states.
 DNA microarray-based approaches for biomarker discovery have been applied
for studying several chronic diseases including diabetes, arthritis, and
cardiovascular disease.
DISEASE RELEVANT BIOMARKERS
41
• Biomarkers for diagnosing SLE and RA, which are chronic autoimmune and
inflammatory disorders, can be screened by expression profiling in leukocytes
because the differential gene expression in leukocytes is clearly relevant to SLE and
RA.
• Osteoarthritis, the degenerative joint disease that can be confused with RA, can also
be diagnosed by using the expression profiling of leukocytes.

42. DETECTION OF CHROMOSOMAL
ABNORMALITIES
There are two very efficient types of microarrays experiment that typically used
for monitoring chromosomal abnormalities:
 Array-based comparative genomic hybridization (aCGH): rapid method to
monitor major DNA copy number changes like deletions or amplifications and
provide more accurate information about chromosomal imbalances.
 SNP-based microarrays (SNP-arrays): SNP arrays contain oligonucleotide
probes spotted systematically to detect the two alleles of a specific SNP locus, in
which both the homozygous and heterozygous genotypes could be detected.
42

43.  SNP arrays have been used is the mapping of human disease susceptibility loci
as published in GWAS. In order to facilitate the GWAS, a detailed human
haplotype map has been created using over a million SNP (The International
HapMap Consortium 2005).
 aCGH provides a lot of information about genomic balance of tumor cells,
mono- or trisomies, amplifications and deletions in a simple experiment. aCGH
technique has been applied to create comprehensive maps of human CNVs.
43
Kang and coworkers (2007) developed a DNA microarray for detecting chromosomal
abnormalities causing various genetic disorders consisting of Down’s syndrome, Patau
syndrome, Edward syndrome, Turner syndrome, Klinefelter syndrome, alpha-thalassemia
retardation-16 and many other diseases.

44. METHYLATION PATTERN ANALYSIS
 Changes in methylation pattern usually generate alterations in gene expression
programs.
 Gene silencing by DNA methylation of specific gene promoters is a well-
known feature of neoplastic cells and plays an important role in normal cell
differentiation and development.
 Tumor cells are generally characterized by the hyper-methylation of tumor
suppressor genes and, in contrast, hypo-methylation of the whole DNA
molecule. 44

45.  In one experiment ten or hundred thousands of distinct and identified
potential methylation sites can be monitored by microarray analysis.
 Analysis of genome wide methylation profile enables to characterize new
tumor classes, or to cluster newly diagnosed cases into already existing groups
based on methylation pattern.
45
Devaney and co-workers (2013) systematically analyzed the epigenetic defects in prostate
cancer (PCa) and tried to find DNA methylation-based biomarkers that may be useful for
the early detection and diagnosis of PCa. They found numerous candidate novel genes
(BNC1, FZD1, RPL39L, SYN2, LMX1B, CXXC5, ZNF783 and CYB5R2) that are
frequently methylated and whose methylation was associated with inactivation of gene
expression in PCa cell lines.

46. GENOMIC ANALYSIS
 Microarray-based genomic DNA profiling (MGDP) technologies are capable
of detecting copy number changes for the whole genome in a single assay.
 Unprecedented sensitivity and cost-effectiveness for a large group of
mutations that have evaded conventional approaches
 Identification of new genes - efficient way of annotating the human genome
and facilitating the use of genomic information for experimental purposes.
46

47. PROTEIN MICROARRAYS
 Protein microarray is an emerging technology that provides a versatile
platform for characterization of hundreds of thousands of proteins in a
highly parallel and high-throughput way.
 Attractive research tools because they enable researchers to study multiple
protein interaction events and activity networks simultaneously.
 They can be used for studying protein–antibody, protein–protein, protein–
substrate, protein–drug, enzyme–substrate or multiple analytes.
47

48. PROTEIN EXPRESSION ARRAYS
Capture arrays/Protein profiling arrays.
 Purpose: Simultaneously quantify the levels of a no. of
proteins produced by a cell, in a highly parallel assay.
 Commonly composed of antibodies immobilized
onto the slide surface (Can be made with MAbs or
PAbs, antibody fragments or synthetic polypeptide
ligands).
 Alternatively the probes used can be peptides,
alternative protein scaffolds or nucleic acid aptamers
(oligonucleotides that bind specifically to proteins).
48

49. 49
PROTEIN EXPRESSION ARRAYS

50. PROTEIN FUNCTION ARRAYS
 Developed for the study and elucidation of the function and interaction of
various biological molecules.
 Constructed using individually purified proteins.
 They enable the study of various biochemical properties of proteins such as:
 Binding activities and structure elucidation of proteins
 Interactions of proteins with proteins, DNA, lipids, drugs, peptides etc.
 Post translational modifications.
 enzyme-substrate relationships
50

51. ARRAY FABRICATION
 Protein arrays have been made of a miniaturised solid support spotted with a
protein, peptide or antibody.
 Immobilization of the proteins to the array surface via chemical (glass arrays),
hydrophobic (hydrogels) or electrostatic (nitrocellulose) interactions.
 Glass arrays are mostly coated with aldehyde or BSA-NHS, which react eith
the primary amines on the N-termini or side chains of proteins.
 Protein and peptide arrays can be spotted/inkjet printed, or peptide arrays can
be synthesized in situ on a solid substrate.
51

52. SAMPLE LABELLING, HYBRIDISATION,
AND DETECTION
 Fluorescence labelling of the proteins or antibodies with Cy3 and Cy5,
FITC(green), Rhodamine(Red) and Texas Red.
 Mono-functional NHS esters of CyDyes can be used to label amine residues
in antibodies and other proteins, or maleimide esters of CyDyes can be used
to label free thiol groups.
 Hybridization conditions for protein arrays are similar to those used in
ELISA assays.
52

53.  The array incubated with the solution containing the protein(s) under
conditions that promote binding between the immobilized probe and the
target molecules.
 Washed under conditions that remove weakly binding target molecules.
 Binding/interactions are then detected and quantified using appropriate
scanners.
 Detection of binding to the array can then be quantified using a standard
DNA microarray scanner.
 Another method of detection is by the use of a secondary labelled antibody
directed to the target protein. 53
SAMPLE LABELLING, HYBRIDISATION,
AND DETECTION

54. APPLICATIONS
 Major areas where protein arrays can potentially be used:
 Proteomics
 Functional analysis
 Diagnostics
54
1. Studying the Pathogenicity of a Disease:
• Proteome arrays could be used to study PTMs, interactions with host proteins,
and enzymatic activity.
• Serum could be taken from individuals and exposed to pathogens and host
antibodies labelled and incubated in the array. Positive signals indicate which
pathogen proteins elicit immune response.
• Useful in vaccine studies.

55. 2. Protein-Protein Interactions
 Proteome arrays have been used to study kinase activities and has been used in serum
profiling.
 Yeast proteome chips have been used to study calmodulin interactions. The study
found 6 known calmodulin binders and 33 additional potential binders. These finding
also led to new consensus motifs in these proteins.
3. Drug Targets
 Entire proteome printed on a chip used to study drug interactions.
4. Autoimmune Diseases
 Charpin et al. used ProtoArray (Invitrogen), a protein Microarray with a
nitrocellulose layer where 8268 human proteins with a GST-Tag are spotted, to detect
specific marker for early stages of RA (WIBG, GABARALP2 und ZNF706).
55

56. 5. Protein–small molecule interaction
 Detection of small molecules in case of drug abuse or doping in sports.
6. Allergen arrays
 Many allergens have the same protein epitopes, which means that the same
sIgE antibody can bind to different allergens from different sources.
 Protein Microarrays or Bead based Systems, can check for cross-reactivity
of allergens in one assay.
 Lower costs, lower sample volume and the possibility of measuring a lot
more components in one experiment.
56
Du et al. developed a Protein Microarray for the detection of different prohibited drugs
in various biological fluids (e.g., whole blood, serum, urine and oral fluid). These arrays
have been used for the detection of different drugs of interest including morphine and
meprobamate in blood.

57. TECHNICAL CHALLENGES TO PROTEIN
MICROARRAYS
 Retaining the secondary and tertiary structure, and therefore activity on the
microarray;
 Identifying and isolating an agent with which to capture the proteins of
interest.
 Accurately measuring the degree of protein binding with a system that is
both sensitive and with a wide dynamic range.
57

58. CARBOHYDRATE/GLYCAN ARRAYS
 Not as widely used as DNA or protein arrays.
 Isolation of carbohydrates from natural sources requires multiple and varied
purification steps.
 Chemical synthesis of carbohydrates/glycans requires multiple selective
protection and deprotection steps during synthesis of oligosaccharides,
because of the often three-dimensional nature of the branched linkages in
oligosaccharides.
 Carbohydrates have a low mass and are mainly hydrophilic, it is difficult to
directly immobilize them onto solid matrices by noncovalent bonding.
58

59.  Immobilization of carbohydrate probes by spotting and immobilization onto
nitrocellulose membranes, by linkage of oligosaccharides to lipids, which
were immobilized to nitrocellulose, or by immobilization to hydrophobic
polystyrene slides.
 Immunostaining of the carbohydrate arrays or detection of binding using
labelled secondary antibodies have been used to detect binding to
carbohydrate arrays.
 Labeling of free carbonyl groups on glycoproteins and carbohydrates can
also be achieved, using CyDye™ hydrazides,
59
CARBOHYDRATE/GLYCAN ARRAYS

60.  Used both as ‘capture’ arrays to detect the expression of a particular
biological molecule such as an antibody, and as assays for functional
interaction studies, such as for studying the specificity of antibodies.
 Thirumalapura and co-workers have constructed lipopolysaccharide (LPS)
microarrays, (LPS from E. coli O111, E. coli O157, and S. typhimurium). They
have used these arrays both to monitor the specificity of MAbs.
 Detection of bacterial toxins using immobilized N-acetyl galactosamine
(GalNAc) and N-acetylneuraminic acid (Neu5Ac) derivatives.
 Use of lectin arrays to profile cell surface carbohydrate expression (mannose,
galactose, and GlcNAc expression) in mammalian cells (Zheng et al., 2005). 60
APPLICATIONS

61. 61

62.  Microarray and its applications: Rajeshwar Govindarajan, Jeyapradha Duraiyan, Karunakaran Kaliyappan,
Murugesan Palanisamy
 Protein microarray technology: Markus F. Templin, Dieter Stoll, Monika Schrenk
 Protein microarray technology: David A Hall, Jason Ptacek.
 DNA microarrays: Types, Applications and their future: Roger Bumgarner
 Applications of DNA Microarray in Disease Diagnostics: Yoo, Seung Min, Jong Hyun Choi, Sang Yup Lee, and
Nae Choon Yoo
 Protein Microarrays: Chien-Sheng Chen Heng Zhu
 Microarray Technology in Practice : Steven Russell, Lisa A. Meadows and Roslin R. Russell
 Microarray Technology Through Applications: Francesco Falciani
 Bioarrays- From Basics to Diagnostics: Krishnarao Appasani
 Microarray Bioinformatics: Dov Stekel
 An Introduction to Microarray Data Analysis: M. Madan Babu
 Microarray technology Ágnes Zvara, Klára Kitajka, Nóra Faragó, László G. Puskás
 DNA Microarray Technology: Devices, Systems, and Applications: Michael J. Heller 62
REFERENCE

63. 63