Outline. Array platform considerations: Comparison between the technologies available in microarrays

Transcription

1 Microarray overview

2 Outline Array platform considerations: Comparison between the technologies available in microarrays Differences in array fabrication Differences in array organization Applications of microarrays in biology What do I need to be aware of when I retrieve public data

3 Typical Experiments Two-color experiment One-color experiment condition control condition control RNA extraction RNA labeling

4 Microarray Nomenclature Sample/Treatment/Condition/Experiment Probes Targets Targets Probes Two-color experiment One-color experiment condition control condition control RNA extraction RNA labeling Targets Probes Slide/Chip Chip

5 DNA Array Technologies (A) Affymetrix (B) Spotting

6 Array fabrication: Affymetrix Chip

7 Array fabrication: Photolithographic Synthesis Lamp Mask Chip

8 Array fabrication: Synthesis of Ordered Oligonucleotide Arrays UV light Light (deprotection) Mask O O O O O Substrate HO HO O O O T T T O O O Light (deprotection) T T O O O Mask Substrate C T T C C O REPEAT CATAT AGCTG TTCCG

9 Visual Inspection

10 Array fabrication: Spotting DNA Material By Robot

11 Array fabrication: Zooming in > 20,000 spots

12 Microarray Imperfections: fabrication and 1. High background hybridization 2. Comets Insufficient blocking Drying during hybridization Inefficient washing Smearing of DNA because of defective UV cross-linking or excessive DNA concentration

13 MicroArray Imperfections : fabrication and printing 7. Irregular spot morphology 8. Donut-shaped spots Due to pin defects: Poor pin tip design or damaged pin tip Irregular distribution of DNA within the spot due to rapid drying of slides.

14 MicroArray Imperfections: fabrication and printing 3. Bubbles during hybridization 4. Particle contamination 5. Pin Blockage Powder, dust or particulates from target

15 Array organization: Array organization: Printed DNA material PCR products of cdna libraries Oligonucleotides of bp-long Number of spots is between 10,000 to 20,000

16 Array organization: GeneChip Expression Array Design Gene Sequence or Expressed Sequence Tag (EST) 5 3 Multiple oligo probes of 25-mer Perfect Match Mismatch

17 Array organization: Affymetrix Expression Arrays Control Induced

18 Affymetrix vs Spotted Spotted array Affymetrix The winner Quantification Probes are in fluidic spots Covalently attached probes Affymetrix Sensitive to Absolute Expression High replicates number within chip Low replicates number within chip Affymetrix Sensitive to Differential expression Twice sample replicates /chip Direct comparison Sample per chip. Spotted Sensitive to Alternative splicing Replicates are of the same fragment Replicates are from different area of the gene Affymetrix Standardization There are no standards All the world use the same chip which is good for comparison Affymetrix/Spotted Economy Customization takes money because of imperfection but some time is the only way Expensive Affymetrix/Spotted Customization Gives flexibility Possible but very expensive Spotted

19 Difference: Analysis flow Quantification Two-color experiment Normalization Pre-processing Filtering noise out Data analysis One-color experiment

20 Common: Expression Experiments set Data Matrix (a result of quantification) Samples Each column comes from one array Genes c Gene expression matrix Each row represent one genes Gene expression levels

21 Applications of microarrays Evolution Most of gene expression differences between chimpanzees human have been detected in their brain. Development Associating apoptosis related gene expression with metamorphosis stages in a variety tissues of Drosophila Behavior Can predict the behavior of honey-bees workers by their brain gene expression. Tissue Molecular signature specific to subtype of cancer tissues Regulation Finding novel regulatory motifs by coupling motif search with co-expression Functional annotation Annotating unknown genes based on co expression (guilty by association) - Question about samples - Question about gene

22 Evolution: : Reveal species specific changes which cannot be revealed by genetic changes Chimanzee and human genomes are 98.7% identical. Therefore phylogentic analysis based on genome sequences cannot separate chimpanzee and human into different clades. Detecting gene expression from three tissues (leukocyte, liver and brain) in these species and orangutan could differentiate human and chimpanzee into mutually exclusive groups. W. Enard et al., Science 296, (2002).

23 Behavior: detecting of small expression differences Before that paper only small amount of genes in brain were shown to have changes in expression which are related to behavior. By conducting the right experimental design using ANOVA in this study they could detect small differences of gene expression in honey bees brain and therefore relate many genes to behavior. By this they could predict the behavior of honey-bees workers by their brain gene expression. C. W. Whitfield, A. M. Cziko, G. E. Robinson, Science 302, (2003).

24 Cancer tissues Finding hundreds of gene that differentiate between two or more groups of cancer tissues Using these set of genes to predict the fate of tissues in a stage where histological and other clinical parameters can not. S. Ramaswamy, K. N. Ross, E. S. Lander, T. R. Golub, Nat Genet 33, (2003).

25 Functional annotation Hybridized mrna from 60 conditions and tissues on predicted exons microarray. Exons that were expressed coherently and mapped adjacently on the chromosome (22) annotated as one gene. D. D. Shoemaker et al., Nature 409, (2001).

26 Regulatory motifs By linkage between regulatory motif and expression profiles, the right combination of regulatory motifs can be found. Y. Pilpel, P. Sudarsanam, G. M. Church, Nat Genet 29, (2001).

27 The Global View of Microarray To be able to predict and classify tissue/species we need many genes To be able to classify genes we need many samples

28 What not to do? Don t t use this technology for detecting the expression of few specific genes.. The reaction of the unchanged genes assumed to be the majority on the chip. It is not recommended to use one way subtraction libraries. The changes on the chip assumed to be symmetric Do not hybridize only two samples. Replicates are needed for noise estimation or reduction.

29 High throughput Annotation As much as the probe sequence is shorter there is a chance that it will catch other genes. Especially genes from the same gene families It would have been very useful to have a database that give you information about the probe specificity. (e.g., GeneAnnote)

30 Check list for working with published data Check whether you are working with absolute expression or with differential expression Check the array annotation annotation is dynamic and may be changed since the time it was entered Try to validate the specificity of the probes. Short probes can catch other gene transcripts with the same domain, gene from the same gene families and may reflect the differences of expression of alternative splice variants. Validate sample annotation!

31 Thank You