1	Microarrays (aka CHIPs) In Vitro and In Silico SoCalBSI July 5, 6, and 7, 2005 Sandra B. Sharp Professor, Biological Sciences, CSULA Sagar Damle Ph.D. Candidate, Division of Biology, Caltech
2	References References for Days 1, 2, and 3 and the optional lecture are found at the end of the optional PowerPoint. They are not in alphabetical order.
3	Objectives for 3 Days To become familiar with the wet-bench basics of 2-color DNA microarrays for expression analysis In the context of 2-color expression arrays, use GeneSpring for selection of differentially expressed genes and for cluster analysis To identify points at which uncertainties and inaccuracies may arise in 2-color expression analysis In the context of 2-color expression arrays, to use EXCEL for processing of primary data Optional To work with annotation of microarray data (optional) To work with a major microarray data base (optional) To become aware of several additional applications for microarray technology (optional) SNP analysis Transcription factor binding site analysis Protein analysis Cell and Tissue analysis
4	Outline for Each Day Day 1 Lecture Principles of expression microarray Workshop Introduction to GeneSpring Day 2 Lecture Principles of major clustering algorithms Workshop Using GeneSpring for cluster analysis Recognizing and dealing with variability in acquired data Day 3 Lecture Sources of uncertainty and erro in microarray analysis Workshop Using EXCEL to process primary data and minimize error Optional Lecture Additional applications of microarray Workshop Annotation, Stanford Microarray Database, Motif Searching (optional)
5	Lecture Outline for Day 1 Post-Genomic Era Analysis of gene expression then and now Principles of microarray (in the context of 2-color arrays) Overview Printing the array Data acquisition Data analysis (filtering and clustering) Making hypotheses about biological function
6	the post-genomic era functional genomics and proteomics Post-genomic = after the sequencing of the genome Functional genomics = the study of the genome in order to determine coordinated expession of mRNAs for all the genes in the genome the control regions of the genes the as yet unknown function of additional expressed but non-coding DNA regions (e.g., template for micro RNAs?)
7	the post-genomic era functional genomics and proteomics Proteomics = the identification of all protein products of each gene and the interactions of all proteins in the cell (leading to systems biology)
8	Expression Microarrays - one approach to functional genomics Expression microarrays are used to determine the identity and relative quantities of each of the different kinds of (m)RNAs expressed by a cell type of interest in a “single experiment”. Examples from your workshops What is expressed in tumor as opposed to normal cells? What is expressed over the period of development of a given tissue or cell type (myogenesis)? AND WHY MIGHT WE CARE?
9	Then and Now Example – Normal vs. tumor tissue THEN AND NOW Hypothesis My favorite mRNA is expressed in tumor tissue, but not in normal tissue. What you need to know The nucleotide sequence of your favorite mRNA How to isolate total (m)RNA from normal and tumor tissues How to perform an mRNA blot hybridization analysis Northern NOW Discovery Which genes are expressed in tumor tissue but not in normal tissue? What you need to know The nucleotide sequence of every possible mRNA How to isolate total mRNA from normal and tumor tissues How to perform and analyze an expression microarray experiment
10	Results - Then and Now Example – Normal vs. tumor tissue
11
12	Expression Microarrays - one approach to functional genomics Microarrays (CHIPs) allow you to study the expression of thousands of genes at once. Each chip has 10,000 - 100,000 distinct spots. Each spot has multiple copies of a sequence that represents a specific gene and thereby its mRNA. Some sequences are known to represent mRNAs from databases of ESTs or from predicted genes, but as yet have no identified function. Some chips “tile” across the entire length of a genome or chromosome, thereby identifying transcripts coded for by template lacking the classical structure of a gene.
13	How big are microarray slides? Remember, 10- to 100,000 genes!
14	Basics of Expression Microarray Technique 5 important steps The Wet Experiment Requires printed slide(s) = CHIPs Requires specialized equipment and software Requires fluorescently labeled sample(s) representing total mRNAs from the sources of interest Requires a hybridization or annealing period Data Acquisition – Where has hybridization occurred? Requires a scanner and associated software Data Processing (Extraction, Storage, and Normalization) Requires scanner software and storage in a tab delimited file Requires EXCEL or similar software Data Analysis and Storage Requires software for cluster and other types of analysis Assigning or hypothesizing regarding function Utilizes cluster analysis together with databases that have annotated genes with known function
15
16
17
18
19	Animation of Expression Dual-Color Microarray Setup http://www.bio.davidson.edu/courses/genomics/chip/chip.html
20	Putting DNA on the Slides Print it onto the slide – PCR and long oligos Synthesize it on the slide by photolithography – somewhat shorter oligos
21
22	http://latin.arizona.edu/~dgalbrai/arrayer.html What information must be built into a program in order to automate printing?
23
24	Data Acquisition - Scanners Make an image of the slide by Exciting the fluorescent dyes Collecting the emitted light Generating digital images of the fluorescent signal and Quantifying the intensity of the signal Two major types Laser excitation with a photomultiplier tube (PMT) detector - faster Filtered white-light excitation with a charge-coupled device (CCD) detector
25
26
27
28
29	Data Storage and Processing Tab Delimited File Output from Scanner Save as EXCEL file Notice the LogRatio column 1 = 2X more mRNA than standard of comparison (red = 2X green) 2 = 4X more -1 = ½ (or 2X less) (red = ½ green) -2 = ¼ Log plot (GeneSpring) or a plot of logs More on Thursday – Lecture and EXCEL workshop (eliminate bad data; correct for background; correct for technical variation within a slide; correct for technical variation between replicate slides; determine the ratio of R to G for each gene)
30	Data Analysis Selecting differentially expressed genes GeneSpring
31	Data Analysis
32	Data analysis (cont’d.) clustering – suggests function Places genes with similar expression patterns in groups. Sometimes genes of unknown function will be grouped with genes of known function. The functions that are known allow the investigator to hypothesize regarding the functions of genes not yet characterized. Examples: Identify genes expressed in a specific tumor type Identify genes coordinately involved in a developmental pathway Identify genes involved in a disease response Identify genes important in cell cycle regulation Identify genes that participate in a biosynthetic pathway Identify genes involved in a drug response
33
34
35	Assigning or Hypothesizes about Biological Meaning of Clusters Identify over-represented functional categories in the clusters (i.e., cluster contains may more genes known to be involved in a specific biological process than would be expected by chance) Requirements for systematic analysis: Controlled vocabulary for describing biological processes (protein biosynthesis\translation, apoptosis\programmed cell death) Standard assignment of genes into functional categories Gene Ontology or GO project at NCBI
36	Gene Ontology (GO) project http://www.geneontology.org/ Purpose: 1) Define controlled terms (ontologies) for description of gene products from 3 aspects: Biological process (DNA repair, mitosis) Molecular function (protein serine/threonine kinase activity, transcription factor activity) Cellular component (nucleus, ribosome) 2) Establish a unified framework for organism-independent gene annotation Characteristics: 1) A gene can have multiple associations in each ontology 2) GO terms are organized in hierarchical structures called directed acyclic graphs (DAGs) - The most general classifications are at top levels of the graph - More specialized classifications at lower levels
37	Hierarchical classification scheme for proteins that function in M-phase of mitosis
38	Online Databases that annotate genes by GO Human Entrez http://www.ncbi.nih.gov/entrez/query.fcgi?db=gene GOA http://www.ebi.ac.uk/GOA/ Mouse – Mouse Genome Informatics (MGI) http://www.informatics.jax.org/ Rat – Rat Genome Database http://rgd.mcw.edu/ Fly – FlyBase http://flybase.bio.indiana.edu/ Arabidopsis – TAIR http://www.arabidopsis.org/ Yeast – Sacchromaces Genome Database http://www.yeastgenome.org/ Affymetrix chips – Netaffx http://www.affymetrix.com
39	Example: Cluster 3, 95 genes
40	Identifying enriched GO categories in clusters In the previous example: Total number of chip’s genes with annotation = 5000 Total number of chip’s genes associated with metabolism GO category = 3,600 (72%) Number of annotated genes in cluster 3 = 73 Number of metabolic genes in cluster 3 = 50 (68%) Is it reasonable to assume that genes in Cluster 3 are enriched for metabolic function? Statistical tests are essential to determine whether enrichment of a certain class of proteins is significant
41	Workshops and Lectures to Come (Day 1) Today How to look at data in GeneSpring Selecting differentially expressed genes GeneSpring demo – yeast cell cycle You are expected to complete the GeneSpring Demo before you leave Day 2 Lecture on clustering – Sagar Workshop - Hands-on with 2 data sets from mammalian cells Cluster analysis – unsupervised methods PCA – principle component analysis k means hierarchical clustering SOMs – self-organizing maps Discrimination – a supervised method ANOVA – analysis of variance Day 3 Lecture – Recognizing shortcomings in microarray data Workshop – Data Refinement
42	Single –color Affymetrix
43
44
45	Affymetrix Chip Scan