FUNCTIONAL GENOMICS
CONTENTS FABRICATION OF CDNA MICROARRAYS NYLON CDNA EXPRESSION ARRAYS PLASTIC MICROARRAYS: A NOVEL ARRAY SUPPORT COMBINING THE BENEFITS OF MACRO- AND MICROARRAYS PREPARING FLUORESCENT PROBES FOR MICROARRAY STUDIES ESCHERICHIA COLI SPOTTED DOUBLE-STRAND DNA MICROARRAYS: RNA EXTRACTION, LABELING, HYBRIDIZATION, QUALITY CONTROL, AND DATA MANAGEMENT ISOLATION OF POLYSOMAL RNA FOR MICROARRAY ANALYSIS PARALLEL ANALYSIS OF GENE COPY NUMBER AND EXPRESSION USING CDNA MICROARRAYS GENOME-WIDE MAPPING OF PROTEIN–DNA INTERACTIONS BY CHROMATIN IMMUNOPRECIPITATION AND DNA MICROARRAY HYBRIDIZATION STATISTICAL ISSUES IN CDNA MICROARRAY DATA ANALYSIS EXPERIMENTAL DESIGN TO MAKE THE MOST OF MICROARRAY STUDIES STATISTICAL METHODS FOR IDENTIFYING DIFFERENTIALLY EXPRESSED GENES IN DNA MICROARRAYS 12 DETECTING STABLE CLUSTERS USING PRINCIPAL COMPONENT ANALYSIS CLUSTERING IN LIFE SCIENCES A PRIMER ON THE VISUALIZATION OF MICROARRAY DATA 15 MICROARRAY DATABASES: STORAGE AND RETRIEVAL OF MICROARRAY DATA FABRICATION OF CDNA MICROARRAYS 1 INTRODUCTION DNA MICROARRAY TECHNOLOGY HAS BEEN USED SUCCESSFULLY TO DETECT THE EXPRESSION OF MANY THOUSANDS OF GENES, TO DETECT DNA POLYMORPHISMS, AND TO MAP GENOMIC DNA CLONES (1–4) IT PERMITS QUANTITATIVE ANALYSIS OF RNAS TRANSCRIBED FROM BOTH KNOWN AND UNKNOWN GENES AND ALLOWS ONE TO COMPARE GENE EXPRESSION PATTERNS IN NORMAL AND PATHOLOGICAL CELLS AND TISSUES (5,6) DNA MICROARRAYS ARE CREATED USING A ROBOT TO SPOT CDNA OR OLIGONUCLEOTIDE SAMPLES ON A SOLID SUBSTRATE, USUALLY A GLASS MICROSCOPE SLIDE, AT HIGH DENSITIES THE SIZES OF SPOTS PRINTED IN DIFFERENT LABORATORIES RANGE FROM 75 TO 150 µM IN DIAMETER THE SPACING BETWEEN SPOTS ON AN ARRAY IS USUALLY 100–200 µM MICROARRAYS WITH AS MANY AS 50,000 SPOTS CAN BE EASILY FABRICATED ON STANDARD 25 MM × 75 MM GLASS MICROSCOPE SLIDES TWO TYPES OF SPOTTED DNA MICROARRAYS ARE IN COMMON USE: CDNA AND SYNTHETIC OLIGONUCLEOTIDE ARRAYS (7,8) THE SURFACE ONTO WHICH THE DNA IS SPOTTED IS CRITICALLY IMPORTANT THE IDEAL SURFACE IMMOBILIZES THE TARGET DNAS, AND IS COMPATIBLE WITH STRINGENT PROBE HYBRIDIZATION AND WASH CONDITIONS (9) GLASS HAS MANY ADVANTAGES AS SUCH A SUPPORT DNA CAN BE COVALENTLY ATTACHED TO TREATED GLASS SURFACES, AND GLASS IS DURABLE ENOUGH TO TOLERATE EXPOSURE TO ELEVATED TEMPERATURES AND HIGH-IONIC-STRENGTH SOLUTIONS IN ADDITION, IT IS NONPOROUS, SO HYBRIDIZATION VOLUMES CAN BE KEPT TO A MINIMUM, ENHANCING THE KINETICS OF ANNEALING PROBES TO TARGETS FINALLY, GLASS ALLOWS PROBES LABELED WITH TWO OR MORE ?UORS TO BE USED, UNLIKE NYLON MEMBRANES, WHICH ARE TYPICALLY PROBED WITH ONE RADIOLABELED PROBE AT A TIME 2 MATERIALS 1 MULTISCREEN ?LTRATION PLATES (MILLIPORE, BEDFORD, MA) 2 QIAGEN QIAPREP 96 TURBO MINIPREP KIT (QIAGEN, VALENCIA, CA) 3 DATP, DGTP, DCTP, AND DTTP (AMERSHAM PHARMACIA, PISCATAWAY, NJ) 4 M13F AND M13R PRIMERS (OPERON, ALAMEDA, CA) 5 TAQ DNA POLYMERASE AND BUFFER (INVITROGEN, CARLSBAD, CA) 6 PCR CYCLEPLATE (ROBBINS, SUNNYVALE, CA) 7 CYCLESEAL POLYMERASE CHAIN REACTION (PCR) PLATE SEALER (ROBBINS) 8 GOLD SEAL MICROSCOPE SLIDES (BECTON DICKINSON, FRANKLIN, NJ) 9 384-WELL PLATES (GENETIX, BOSTON, MA) 10 SUCCINIC ANHYDRIDE (SIGMA, ST LOUIS, MO) IN 325 ML OF 1-METHY-2-PYRROLIDINONE (SIGMA) 3 METHODS 3 1 SELECTION AND PREPARATION OF CDNA CLONES 3 1 1 SELECTION OF CLONES MICROARRAYS ARE USUALLY MADE WITH DNA FRAGMENTS THAT HAVE BEEN AMPLI?ED BY PCR FROM PLASMID SAMPLES OR DIRECTLY FROM CHROMOSOMAL DNA THE SIZES OF THE PCR PRODUCTS ON OUR ARRAYS RANGE FROM 0 5 TO 2 KB THEY ATTACH WELL TO THE GLASS SURFACE THE AMOUNT OF DNA DEPOSITED PER SPOT DEPENDS ON THE PINS CHOSEN FOR PRINTING, BUT ELEMENTS WITH 250 PG TO 1 NG OF DNA (UP TO 9 × 108 MOLECULES) GIVE AMPLE SIGNALS MANY OF THE CDNA CLONES THAT HAVE BEEN ARRAYED BY LABORATORIES IN THE PUBLIC DOMAIN HAVE COME FROM THE INTEGRATED MOLECULAR ANALYSIS OF GENOMES AND EXPRESSION (IMAGE) CONSORTIUM SET FIVE MILLION HUMAN IMAGE CLONES HAVE BEEN COLLECTED AND ARE AVAILABLE FROM INVITROGEN/RESEARCH GENETICS (WWW RESGEN COM/PRODUCTS/IMAGECLONES PHP3) SEQUENCE-VERI?ED CDNA CLONES FROM HUMANS, MICE, AND RATS ARE ALSO AVAILABLE FROM INVITROGEN/RESEARCH GENETICS CDNA CLONES CAN ALSO BE OBTAINED FROM OTHER SOURCES THE 15,000 NATIONAL INSTITUTE OF AGING (NIA) MOUSE CDNA SET HAS BEEN DISTRIBUTED TO MANY ACA- DEMIC CENTERS (HTTP://LGSUN GRC NIA NIH GOV/CDNA/15K/HSC HTML) OTHER MOUSE CDNA COLLECTIONS INCLUDE THE BRAIN MOLECULAR ANATOMY PROJECT (BMAP) (HTTP://BRAINEST ENG UIOWA EDU), AND RIKEN (HTTP://GENOME RTC RIKEN GO JP) CLONE SETS IN PREPARING OUR ARRAYS, WE HAVE USED THE NIA AND BMAP COLLEC- TIONS AND ARE IN THE PROCESS OF SEQUENCING THE 5V ENDS OF THE 41,000 CLONES IN THE COMBINED SET IN COLLABORATION WITH SCIENTISTS AT THE KOREA RESEARCH INSTITUTE OF BIOSCIENCE AND BIOTECHNOLOGY NOTE THAT MOST CDNA COLLECTIONS SUFFER FROM SOME GRIDDING ERRORS AND WELL-TO-WELL CROSS CONTAMINATION PRIMERS THE SEQUENCES OF THE FORWARD (M13F) AND REVERSE (M13R) PRIMERS USED ARE 5V-GTTGTAAAACGACGGCCAGTG-3V AND 5V-CACACAGGAAA CAGCTATG-3V, RESPECTIVELY A VARIETY OF METHODS ARE AVAILABLE FOR PURIFYING CDNA SAMPLES WE USE QIAPREP 96 TURBO MINIPREP KITS AND A QIAGEN BIOROBOT 8000 (QIAGEN) FOR PLASMID ISOLATIONS BUT CHEAPER, SEMIAUTOMATED TECHNIQUES CAN BE USED AS WELL WE PCR DNAS WITH A TETRAD MULTICYCLER (MJ RESEARCH, INCLINE VILLAGE, NV) AND PURIFY THE PRODUCTS WITH MULTISCREEN ?LTRATION PLATES (MILLIPORE) 3 1 3 PURI? CATION OF PLASMID 1 CULTURE THE BACTERIAL CLONES OVERNIGHT IN 1 3 ML OF LURIA–BERTANI (LB) MEDIUM CONTAINING 100 µG/ML OF CARBENICILLIN AT 37°C, SHAKING THEM AT 300 RPM IN 96-WELL ?AT-BOTTOMED BLOCKS 2 HARVEST THE BACTERIA BY CENTRIFUGING THE BLOCKS FOR 5 MIN AT 1500G IN AN EPPENDORF CENTRIFUGE 5810R (EPPENDORF, WESTBURY, NY) REMOVE THE LB BY INVERTING THE BLOCK THE CELL PELLETS CAN BE STORED AT –20°C 3 PREPARE CDNA USING THE BIOROBOT 8000, OR FOLLOW THE QIAGEN QIAPREP 96 TURBO MINIPREP KIT PROTOCOL FOR MANUAL EXTRACTION 4 ELUTE THE DNA WITH 100 µL OF BUFFER EB (10 MM TRIS-HCL, PH 8 5) INCLUDED IN THE QIAPREP 96 TURBO MINIPREP KIT THE PLASMID DNA YIELD SHOULD BE 5–10 µG PER PREP 3 1 4 PCR AMPLI? CATION 1 DILUTE THE PLASMID SOLUTION 110 WITH 1X TE (10 MM TRIS-HCL, PH 8 0, 1 MM EDTA) 2 FOR EACH 96-WELL PLATE TO BE AMPLI?ED, PREPARE A PCR REACTION MIXTURE CONTAINING THE FOLLOWING INGREDIENTS: 1000 µL OF 10X PCR BUFFER (INVITROGEN), 20 µL EACH OF DATP, DGTP, DCTP, AND DTTP (100 MM EACH; AMERSHAM PHARMACIA), 5 µL EACH OF M13F AND M13R (1 MM EACH; OPERON), 100 µL OF TAQ DNA POLYMERASE (5 U/µL; INVITROGEN), AND 8800 µL OF DDH2O 3 ADD 100 µL OF PCR REACTION MIX TO EACH WELL OF A PCR CYCLEPLATE (ROBBINS) PLUS 5 µL OF DILUTED PLASMID TEMPLATE SEAL THE WELLS WITH CYCLESEAL PCR PLATE SEALER (ROBBINS) (PREPARE TWO PLATES FOR AMPLI?CATION FROM EACH ORIGINAL SOURCE PLATE TO GIVE A ?NAL VOLUME OF 200 µL OF EACH PRODUCT ) 4 USE THE FOLLOWING PCR CONDITIONS: 96°C FOR 2 MIN; 30 CYCLES AT 94°C FOR 30 S, 55°C FOR 30 S, 72°C FOR 1 MIN 30 S; 72°C FOR 5 MIN; AND COOL TO AMBIENT TEMPERATURE FUNCTIONAL GENOMICS 5 ANALYZE 2 µL OF EACH PRODUCT ON 2% AGAROSE GELS WE USE AN OWL MILLIPEDE A6 GEL SYSTEM (PORTSMOUTH, NH) WITH EIGHT 50-TOOTH COMBS THIS ALLOWS US TO RUN 384 SAMPLES PER GEL 3 1 5 CLEANUP OF PCR PRODUCT 1 TRANSFER THE PCR PRODUCTS FROM THE TWO DUPLICATE PCR CYCLEPATES TO ONE MILLIPORE MULTISCREEN PCR PLATE USING THE QIAGEN BIOROBOT 8000 2 PLACE THE MULTISCREEN PLATE ON A VACUUM MANIFOLD APPLY THE VACUUM TO DRY THE PLATE 3 ADD 100 µL OF DDH2O TO EACH WELL 4 SHAKE THE PLATE FOR 30 MIN AT 300 RPM 5 TRANSFER THE PURI?ED PCR PRODUCTS TO A 96-WELL PLATE 6 STORE THE PCR PRODUCTS IN A –20°C FREEZER 3 2 CREATING CDNA MICROARRAYS (SEE NOTE 1) ROBOTS ARE ROUTINELY USED TO APPLY DNA SAMPLES TO GLASS MICROSCOPE SLIDES THE SLIDES ARE TREATED WITH POLY-L-LYSINE OR OTHER CHEMICAL COATINGS SOME INVESTIGATORS IRRADIATE THE PRINTED ARRAYS WITH UV LIGHT SLIDES COATED WITH POLY-L-LYSINE HAVE A POSITIVELY CHARGED SURFACE, HOWEVER, AND THE NEGATIVELY CHARGED DNA MOLECULES BIND QUITE TIGHTLY WITHOUT CROSSLINKING FINALLY, THE HYDROPHOBIC CHARACTER OF THE GLASS SURFACE MINIMIZES SPREADING OF THE PRINTED SPOTS POLY-L-LYSINE-COATED SLIDES ARE INEXPENSIVE TO MAKE, AND WE HAVE FOUND THAT THEY WORK QUITE WELL ABOUT 1 NL OF PCR PRODUCT IS SPOTTED PER ELEMENT MANY PRINTERS ARE COMMERCIALLY AVAILABLE ALTERNATIVELY, ONE CAN BE BUILT IN-HOUSE (FOR DETAILED INSTRUCTIONS, VISIT HTTP://CMGM STANFORD EDU/PBROWN/MGUIDE/INDEX HTML) AFTER THE ARRAYS ARE PRINTED, RESIDUAL AMINES ARE BLOCKED WITH SUCCINIC ANHYDRIDE (SEE HTTP://CMGM STANFORD EDU/PBROWN/MGUIDE/INDEX HTML) 3 2 1 COATING SLIDES WITH POLY-L-LYSINE 1 PREPARE CLEANING SOLUTION BY DISSOLVING 100 G OF NAOH IN 400 ML OF DDH2O ADD 600 ML OF ABSOLUTE ETHANOL AND STIR UNTIL THE SOLUTION CLEARS 2 PLACE GOLD SEAL MICROSCOPE SLIDES (BECTON DICKINSON) INTO 30 STAINLESS-STEEL SLIDE RACKS (WHEATON, MILLVILLE, NJ) PLACE THE RACKS IN A GLASS TANK WITH 500 ML OF CLEANING SOLUTION WORK WITH FOUR RACKS (120 SLIDES IN TOTAL) AT A TIME 3 SHAKE AT 60 RPM FOR 2 H 4 WASH WITH DDH2O FOUR TIMES, 3 MIN FOR EACH WASH 5 MAKE A POLY-L-LYSINE SOLUTION BY MIXING 80 ML OF 0 1% (W/V) POLY-L-LYSINE WITH 6 TRANSFER TWO RACKS INTO ONE PLASTIC TRAY WITH 400 ML OF COATING SOLUTION 7 SHAKE AT 60 RPM FOR 1 H FUNCTIONAL GENOMICS 8 RINSE THE SLIDES THREE TIMES WITH DDH2O 9 DRY THE SLIDES BY PLACING THEM IN RACKS (SHANDON LIPSHAW, PITTSBURGH, PA) AND SPINNING THEM AT 130G FOR 5 MIN IN A SORVALL SUPER T21 CENTRIFUGE WITH AN ST-H750 SWINGING BUCKET ROTOR PLACE ONE SLIDE RACK IN EACH BUCKET 10 STORE THE SLIDES IN PLASTIC STORAGE BOXES AND AGE THEM FOR 2 WK BEFORE PRINTING DNA ON THEM 3 2 2 SPOTTING DNA ON COATED SLIDES WE USE THE FOLLOWING PARAMETERS TO PRINT 11,136 ELEMENT ARRAYS WITH AN OMNIGRID ROBOT HAVING A SERVER ARM (GENEMACHINES, SAN CARLOS, CA): 4 × 4 SMP3 PINS (TELECHEM, SUNNYVALE, CA), 160 × 160 µM SPACING, 27 × 26 SPOTS IN EACH SUBARRAY, SINGLE DOT PER SAMPLE WE USE THE FOLLOWING PRINTING PARAMETERS: VELOCITY OF 13 75 CM/S, ACCELERATION OF 20 CM/S2, DECELERA- TION OF 20 CM/S2 WE PRINT TWO IDENTICAL ARRAYS ON EACH SLIDE 1 ADJUST THE RELATIVE HUMIDITY OF THE ARRAYER CHAMBER TO 45–55% AND THE TEMPERA- TURE TO 22°C 2 DILUTE THE PURI?ED PCR PRODUCTS 11 WITH DIMETHYLSULFOXIDE (DMSO) (SIGMA) (SEE NOTE 2) TRANSFER 10-µL ALIQUOTS OF THE SAMPLES TO GENETIX 384-WELL PLATES (GENETIX) 3 LOAD THE PLATES INTO THE CASSETTE OF THE SERVER ARM THREE SUCH CASSETTES HOLD 36 PLATES RELOAD THE CASSETTES IN MIDRUN IF MORE THAN 36 PLATES OF SAMPLES ARE TO BE PRINTED IT TAKES ABOUT 24 H TO PRINT 100 SLIDES WITH 2 × 11,136 ELEMENTS ON THEM 4 LABEL THE SLIDES EXAMINE THE ?RST SLIDE IN THE SERIES UNDER A MICROSCOPE MARK THE FOUR CORNERS OF THE ARRAY (OR THE SEPARATE ARRAYS IF THERE ARE MORE THAN ONE ON THE SLIDE) WITH A SCRIBE USE THIS INDEXED SLIDE TO DRAW A TEMPLATE ON A SECOND MICROSCOPE SLIDE SHOWING WHERE THE COVER SLIP SHOULD BE PLACED DURING THE HYBRIDIZATION STEP REMOVE THE REMAINING SLIDES FROM THE ARRAYER AND STORE THEM IN A PLASTIC BOX 3 2 3 POSTPROCESSING WE OFTEN POSTPROCESS OUR ARRAYS AFTER STORING THEM FOR SEVERAL DAYS THIS MAY NOT BE NECESSARY AS OTHERS HAVE ARGUED, BUT IT IS SOMETIMES CONVENIENT MANY WORKERS RECOMMEND UV CROSSLINKING THE DNA TO THE SLIDE SURFACE BY EXPOSING THE ARRAYS TO 450 MJ OF UV IRRADIATION IN A STRATALINKER (STRATAGENE, LA JOLLA, CA) AS NOTED, THIS STEP IS OPTIONAL, AND WE HAVE NOT FOUND IT TO BE CRITICAL 1 INSERT 30 SLIDES INTO A STAINLESS STEEL RACK AND PLACE EACH RACK IN A SMALL GLASS 2 IN A CHEMICAL FUME HOOD, DISSOLVE 6 G OF SUCCINIC ANHYDRIDE (SIGMA) IN 325 ML OF 1-METHY-2-PYRROLIDINONE (SIGMA) IN A GLASS BEAKER BY STIRRING FUNCTIONAL GENOMICS 3 ADD 25 ML OF 1 M SODIUM BORATE BUFFER (PH 8 0) TO THE BEAKER AS SOON AS THE SUCCINIC ANHYDRIDE IS DISSOLVED 4 RAPIDLY POUR THE SOLUTION INTO THE GLASS TANK 5 PLACE THE GLASS TANK ON A PLATFORM SHAKER AND SHAKE AT 60 RPM FOR 20 MIN IN THE HOOD WHILE THE SLIDES ARE INCUBATING ON THE SHAKER, PREPARE A BOILING WATER BATH 6 TRANSFER THE SLIDES TO A CONTAINER WITH 0 1% SODIUM DODECYL SULFATE SOLUTION SHAKE AT 60 RPM FOR 3 MIN 7 WASH THE SLIDES WITH DDH2O FOR 2 MIN REPEAT THE WASH TWO MORE TIMES 8 PLACE THE SLIDES IN THE BOILING WATER BATH TURN OFF THE HEAT IMMEDIATELY AFTER SUBMERGING THE SLIDES IN THE WATER DENATURE THE DNA FOR 2 MIN IN THE WATER BATH 9 TRANSFER THE SLIDES TO A CONTAINER WITH 100% ETHANOL AND INCUBATE FOR 4 MIN 10 DRY THE SLIDES IN A CENTRIFUGE AT 130G FOR 5 MIN (SEE SUBHEADING 3 2 1 , STEP 9) AND STORE THEM IN A CLEAN PLASTIC BOX THE SLIDES ARE NOW READY TO BE PROBED (SEE NOTE 3) 4 NOTES 1 THE METHODS FOR PRINTING SLIDES DESCRIBED IN THIS CHAPTER ARE SOMEWHAT TEDIOUS, BUT THEY ARE ROBUST AND INEXPENSIVE 2 WE RECOMMEND DISSOLVING THE DNAS TO BE PRINTED IN 50% DMSO INSTEAD OF AQUEOUS BUFFERS BECAUSE THIS IS A SIMPLE SOLUTION TO THE PROBLEM OF SAMPLE EVAPORATION DURING LONG PRINTING RUNS (10) 3 THE PROBE-LABELING TECHNIQUE THAT WE DESCRIBE IN CHAPTER 4 WORKS WELL WITH SLIDES PREPARED ACCORDING TO THE PROTOCOLS WE HAVE GIVEN REFERENCES 1 SCHENA, M , SHALON, D , DAVIS, R W , AND BROWN, P O (1995) QUANTITATIVE MONITORING OF GENE EXPRESSION PATTERNS WITH A COMPLEMENTARY DNA MICROARRAY SCIENCE 270, 467–470 2 SCHENA, M , SHALON, D , HELLER, R , CHAI, A , BROWN, P O , AND DAVIS, R W (1996) PARALLEL HUMAN GENOME ANALYSIS: MICROARRAY-BASED EXPRESSION MONITORING OF GENES PROC NATL ACAD SCI USA 93, 10,614–10,619 3 DERISI, J , VISHWANATH, R L , AND BROWN, P O (1997) EXPLORING THE METABOLIC AND GENETIC CONTROL OF GENE EXPRESSION ON A GENOMIC SCALE SCIENCE 278, 680–686 4 SAPOLSKY, R J AND LIPSHUTZ, R J (1996) MAPPING GENOMIC LIBRARY CLONES USING OLIGONUCLEOTIDE ARRAYS GENOMICS 33, 445–456 5 DERISI, J , PENLAND, L , BROWN, P O , BITTNER, M L , MELTZER, P S , RAY, M , CHEN, Y , SU, Y A , AND TRENT, J M (1996) USE OF A CDNA MICROARRAY TO ANALYSE GENE EXPRESSION PATTERNS IN HUMAN CANCER NAT GENET 14, 457–460 6 HELLER, R A , SCHENA, M , CHAI, A , SHALON, D , BEDILION, T , GILMORE, J , WOOL- LEY, D E , AND DAVIS, R W (1997) DISCOVERY AND ANALYSIS OF IN?AMMATORY DISEASE-RELATED GENES USING CDNA MICROARRARY PROC NATL ACAD SCI USA 94, 2150–2155 FUNCTIONAL GENOMICS 7 SHALON, D , SMITH, S J , AND BROWN, P O (1996) A DNA MICROARRAY SYSTEM FOR ANALYZING COMPLEX DNA SAMPLES USING TWO-COLOR ?UORESCENT PROBE HYBRIDIZATION GENOME RES 6, 639–645 8 LIPSHUTZ, R J , FODOR, S P A , GINGERAS, T R , AND LOCKHART, D J (1999) HIGH DENSITY SYNTHETIC OLIGONUCLEOTIDE ARRAYS NAT GENET 21(SUPPL ), 20–24 9 CHEUNG, V G , MORLEY, M , AGUILAR, F , MASSIMI, A , KUCHERLAPATI, R , AND CHILDS, G (1999) MAKING AND READING MICROARRAYS NAT GENET 21(SUPPL ), 15–19 10 HEGDE, P , QI, R , ABERNATHY, K , GAY, C , DHARAP, S , GASPARD, R , HUGHES, J E , SNESRUD, E , LEE, N , AND QUACKENBUSH, J (2000) A CONCISE GUIDE TO CDNA MICROARRAY ANALYSIS BIOTECHNIQUES 29, 548–556