Class Note for BME 510 at UA 5
Popular in Course
Popular in Department
This 5 page Class Notes was uploaded by an elite notetaker on Friday February 6, 2015. The Class Notes belongs to a course at University of Arizona taught by a professor in Fall. Since its upload, it has received 12 views.
Reviews for Class Note for BME 510 at UA 5
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 02/06/15
a 1999 Nature America Inc httpgeneticsnaturecom review 1999 Nature America Inc httpgeneticsnaturecom Expression profiling using cDNA microarrays David J Duggan Michael Bittner Yidong Chen Paul Meltzer amp Jeffrey M Trent Cancer Genetics Branch Natiunal Human Genume Research Institute Na tiunal Institutes ufHealth Bethesda Maryland 20892 USA eemail jtrentnhg7inihg0v cDNA microarrays are capable of profiling gene expression patterns of tens of thousands of genes in a single experiment DNA targets in the form of 3 expressed sequence tags ESTs are arrayed onto glass slides or membranes and probed with fluorescent or radioactivelylabelled cDNAs Here we review technical aspects of cDNA microarrays including the general principles fabrication of the arrays target labelling image analysis and data extraction management and mining Ambitious projects aimed at cloning mapping and sequencing the genomes of various organisms including that of Homo sapz ens have been launched worldwide In all cases the fruits of these labours will provide a solid platform from which to attempt the larger goal of understanding how genomes result in the organisms they specify The success of these international efforts is imprese sive So far complete genomic sequences of 17 organisms includ ing the eukaryote Saccharomyces cerevisiae have been produced The mapping both genetic and physical and sequencing phases of the Human Genome Project are ahead of schedule Researchers have catalogued more than 11 million expressed sequence tagged sites ESTs corresponding with 52907 unique human genes1 wwwaninlmnihgovUniGene However the function ex pression and regulation of more than 80 of them has yet to be fathomed The next phase of the human genome project will place strong emphasis on assigning function to these genes The ability to identify genes at the nucleic acid level rather than proceeding from a known protein to its chromosomal counter part has prompted efforts to likewise extract functional informa tion at the nucleic acid level Two methods are currently in use The sequence approach has led to the discovery of a wide Variety of sequence motifs encoding structural domains such as DNA binding and nucleotideebinding domainsz thus providing clues to gene function Another route for exploring the function of a gene is by determining its pattern of expression The accumula tion of expression data has yet to reach the point at which it is possible to speak of expression motifs but it does suggest that this is a plausible outcome of the approach3 5 Various methods are available for detecting and quantitating gene expression levels including northern blots SI nuclease prof tection7 differential display8 sequencing of cDNA libraries9 10 and serial analysis of gene expression11 SAGE Augmenting this coterie are two arrayebased technologiesicDNA and oligonue cleotide arrays These allow one to study expression levels in par allelmz m thus providing static information about gene expression that is in which tissues the gene is expressed and dynamic information that is how the expression pattern of one gene relates to those of others The high degree of digital data extraction and processing of these techniques supports a Variety of samples or experimental conditions Although both cDNA and oligonucleotide arrays are capable of analysing patterns of gene expression fundamental differences exist between the methods Here we focus primarily on technical aspects of cDNA microarrays although some comparison with 10 the oligonucleotide array see page 20 of this issue ref 14 will be made where appropriate Principle of method As reviewed by Ed Southern on page 5 of this issue hybridization between nucleic acids one of which is immobilized on a matrix provides a core capability of molecular biology This method provides high sensitivity and speci city of detection as a conse quence of exquisite mutual selectivity between complementary strands of nucleic acids Historically most applications of this method have employed a single pure labelled oligonucleotide or polynucleotide species in the liquid phase and complex mixtures of polynucleotides attached to a solid support Transcript abun dance is assayed by immobilizing mRNA or total RNA elece trophoretically separated or in bulk on membranes and then incubating with a radioactively labelled geneespecific target If multiple RNA samples are immobilized on the same matrix one obtains information about the quantity of a particular message present in each RNA pool cDNA arrays alter this strategy in several ways Fig 1 In an array experiment many geneespecific polynucleotides derived from the 3 end of RNA transcripts are individually arrayed on a single matrix This matrix is then simultaneously probed with uorescently tagged cDNA representations of total RNA pools from test and reference cells allowing one to determine the rela tive amount of transcript present in the pool by the type of uorescent signal generated Relative message abundance is inherently based on a direct comparison between a test cell state and a reference cell state an internal control is thus provided for each measurement Fig 2 The scheme is similar when using radiolabelled probe but it is not possible to carry out simultanee ous hybridization of test and reference samples In such cases serial or parallel hybridization is required introducing the possi bility of higher variability in comparisons of expression level The adaptable nature of the fabrication and hybridization methods allows the technique to be applied widelyithe only limitations are the availability of clones for the solid phase and the quality of RNA samples derived from the cells or tissues to be compared This is illustrated by diverse applications that include investigating gene expression in the roots and leaves of Arabidopsis thaliana3 human T cells exposed to phorbol ester rheumatoid arthritis and in ammatory bowel disease16tumorie genic versus nonetumorigenic cell lines4 the diauxic shift from anaerobic to aerobic metabolism in S cerevisiae5 17 yeast nature genetics supplement 0 volume 21 o january 1999 9 1999 Nature America Inc gt hnpdlgeneiicsnaturecom 95 1999 Nature Ameritz Inc gt httpIgenetianatureoorn review Fly 1 thA mltioanay hema Template fol gene of lnteiert ale obtalned and amplified by m Following bunntauon and quality control aliquot 5 nl ale printed on coated glau micro ope llde urlng a omputemonvolled nlgn peed iobot Total RNA on both the text and reference ample u rluoiertently labelled With either Lye oi yesguw urlng a lngle iound of YeVeYXe tianrtnbuon The tluoiertent taiget ale pooled and allowed to nybndlze undei Ulngent ondltlon to the lone on the anay larei exa r tlonofthelnltoipoialedtaigelsyleldxanemlxxlon wtth a Havatevl rbettia wnltn u meaxuied urlng a arlrllrlg tontotal larei mltiortobe Monochrome lmage iron the rtannei ale imported into oftwaie in which the image ale pxeudormlouiedand meiged lnfoimanon about the lone lrldudlrlg gene name tlone ldentlflev lntenrlty valuex lntenrlty iatlor normalization onxtant and confidence lrlteivalx u attached to each target bata non a lngle hybildlzatlorl expeilment u Vlewed a a noimallzed rate that u nyezzyesl in which rlgnlntantdewatlonr non l no change ale lrldKathe of lntieared gtl oi detieared lt0 level of gene exbierrlon ielatlve to the reference ample ln addluon dab non multlple expeilmerlt an be examined urlng any numbeiotdan mlnlngtool DNA lone tobotle pnntln g murlne T cells challenged with AsphorbolrllrmyrlstateVISV acetate and ln Stveptammuspneumamaem Fabrication Productlon of arrays beglns wlth the selectlon ofthe probes to be prlnted on the array In many cases these are chosen dlrectly from databases including GenBank ref 19L dbEST ref 20 and Uanene ref DI the resource backbones ofthe array technolor gles see page 25 ofthls lssue ref 21 Additionally tulllength cDNAsl collectlons of partially sequenced cDNAs or ESTsl or randomly chosen cDNAs from any library oflnterest can be used Arrays for higher eukaryotes are typically based on the EST pore tlons ofthese prolectsl whereas for yeast and prokaryotesl probes are usually generated by amplifying genomic DNA with genes speci c primers Given the expense ofobtalnlng clones produce lng DNA from them and prlntlng theml lt ls usually preferable to produce arrays wlth a low redundancy of representation so as to survey the broadest posslble set ofgenes In thls regard the human Uanene database represents an excellent model ofthe kind oflnformatlonal base one needs both to choose clones and to evaluate expresslon pro les It lncludes a summary oflnformatlon about the functlon ofa partlcular genel lts genomlc locatlonl clones that contaln the gene and connecr tlons to other relevant databases and llterature sources On the other hand no other organisms have such a wellrdeveloped EST database a llmltatlonl glven that cDNA mlcroarrays also permlt the assay ofuncharacterlzed cDNAs whlch may represent genes wlth lnformatlve expresslon patterns cDNA arrays are produced byspottlngPCR products ofapproxr lmately 06714 kb representlng specl c genes onto a mall le These are usually generated from purl ed templatesl so that cellular oonr tamlnants do not nd thelr way onto the array Typlcally the PCR product ls partlally purl ed by preclpltauom gelr ltrauonl or both ito remove unwanted saltsl detergents PCR prlmers and protelns present ln the PCR cocktall For both glass and membrane matnr cesl each array element ls generated by the deposition of a few nanollters ofpurl edPCR producb typically otloosoougml see page 18 ofthls lssue ref 22 Prlntlng ls carrled out by a robot that spots a sample ofeach gene product onto a number ofmatrlces ln a serlal operatlon The rst spottlng robots relled on contact prlntlng wlth a devlce not unllke a fountaln pen Many varlatlons on thls design are now available see page 31 of thls lssue ref 21 in nature genelkssupplemenl volume 21 january 1999 pcu ambllneatlon bunneatlon AMP text teteten e exltltztlorl NegR laxeil laretz iSerxe tiznmlptlon label with uoidye s 1 3amp1 omputei nybnulzetamet analyxl to m letoanay addltlon to a spotter that ls essentlally a caplllary tubal to whlch a low but constant pressure ls applleds None contact prlntlng modesl uslng elther plezo or mkrjet devlcesl are also belng evaluated The types of membranes commonly used are nltrocellulose and charged nylon commerclal varieties that are used for Vane ous blottlng assays Glassrbased arrays are most often made on microscope slides which have low inherent uorescence These are coated wlth polyrlyslnel amlno sllanes or amlnorreactwe sllaneer which enhance both the hydrophoblclty of the sllde and the adherence of the deposited DNAV They also limit the spread ofthe spotted DNA droplet on the sllde In most cases DNA ls crossrllnked to the matrlx by ultravlo let lrradlatlon After XaUOHl resldual amlnes on the sllde surface are reacted wlth succlnlc anhydrlde to reduce the posltlve charge at the surface As a nal step some percentage of the DNA deposlted ls rendered slnglerstranded by heat or alkali see page 19 ofthls lssue for a detalled description of procedureszl The state of bound DNA ls lllrde neds It ls deposited in double stranded forml intrarstrand crossrllnked to some extentl and may well have multlple constralnlng contacts wlth the matrlx along lts length lnduced by drylng the DNA onto the matrlx Figs 3 It ls therefore probably not the best hybrldlzatlon probe One can lmaglne that ollgonucleotlde matricesl wlth thelr short chalns and slngle polnts ofconstralnt at each chaln endl may well be a far more accessible probe for hybridization Against thls advantagel howeverl must be welghed the dlsadvantages of uslng shortrchaln detectors Chlef among these are the varlatlons ln meltlng temperature due to ATrGC COmpOSlUOnl and the reduce tlon ln specl clty due to truncatlng the number ofnucleotldes from hundreds to as few as twenty A format ln whlch the accesslr blllty ofa slmply tethered slnglerstranded probe could be come blned with the speci city of a long probe would provide a considerable improvement for the elds Target labelling and hybridization The targets for arrays are labelled representations of cellul r mRNA pools Typlcally reverse transcription from an ollgordT prlmer ls used This has the virtue ofproduclng a labelled product from the 3 end ofthe genel dlrectly complementary to Immoblr llzed targets syntheslzed from ESTs Frequently total RNA pools rather than mRNA selected on ollgordT are labelled to maxlr mlze the amount of message that can be obtalned from a glven 11 1999 Nature America Inc httpgeneticsnaturecom review 1999 Nature America Inc httpgeneticsnaturecom CI eoooo Signal intensities across array probes I Cye5 Signal E eyes signa E i l I 57 CDKN1A Probes i 39n r i I I I 2 3 i E i a a g ii I In 99 iquot I 8 ii i a Ii 391 i it 394 39MAU Pixels 8 8 Irradiated a g 39v t control 393 v lquot CDKN1A probe Fig 2 Quantitations from two colour hybridization a A segment of an array to which targets from y irradiated ML1 cells red and untreated ML1 cells green are hybridized Highly differential hybridization is visible at the detectors for CDKN1A and MYC boxed b Intensity along a horizontal axis running through CDKN1A and several detectors on either side The intensity profiles are nearly coincident at each probe except CDKN1A At CDKN1A the signal from the unin duced cells is near the threshold of detection whereas the signal from the induced cells is considerably greater amount of tissue The purity of RNA is a critical factor in hybridization performance particularly when using uorescence as cellular protein lipid and carbohydrate can mediate signi cant nonspeci c binding of uorescently labelled cDNAs to slide sur faces For radioactive detection 33F dCTP is preferred to more energetic emitters as array elements are physically close to each other and strong hybridization with a radioactive target can easily interfere with detection of weak hybridization in surrounding targets As uorescent labels Cye3dUTP and CyeSdUTP are fre quently paired as they have relatively high incorporation ef cien cies with reverse transcriptase good photostability and yield and are widely separated in their excitation and emission spectra allowing highly discriminating optical ltration A clear limitation to the application of this technology is the large amount of RNA required per hybridization For adequate uorescence the total RNA required per target per array is 50 200 pg 2 5 pg are required when using polyA mRNA For mRNA present as a single transcript per cell approximately 1 transcript per 100000 application of target derived from 100 pg of total RNA over an 800 mm2 hybridization area containing 200pm diameter probes will result in approximately 300 tran scripts being suf ciently close to the target to have a chance to hybridize Thus if the uorescently tagged transcripts are on average 600 bp have an average of 2 uor tags per 100 bp and hybridize all of them to their probe approximately 12 uors will be present in a 100pm2 scanned pixel from that probe Such low levels of signal are at the lower limit of uorescence detec tion and could easily be rendered undetectable by assay noise Although radioactive targets may have a higher intrinsic detectability they too reach a level of dilution that prohibits effective detection thus precluding experimentation on very small numbers of cells Fig 4 A variety of means by which to improve signal from limited RNA has been proposed These are being evaluated by our labo ratory and many others Ef cient mixing of the hybridization uid should bring more molecules into contact with their cog nate probe increasing the number of productive events This entails however a larger mixing volume which might offset the potential gain Methods that produce multiple copies of mRNA using highly ef cient phage RNA polymerases have been 12 developed A version of this approach in which labelled target cRNA is made directly from a cDNA pool having a T7 RNA polymerase promoter site at one end via in vitro transcription has been applied to arrays Posthybridization ampli cation methods have also been reported in which detectable molecules are precipitated at the target by the action of enzymes sand wiched to the cDNA target24 Detection of hybridized species using mass spectroscopy or local changes in electronic proper ties can also be imagined25 26 Image analysis and data extraction The highly regular arrangement of detector elements and crisply delineated signals that result from robotic printing and confocal imaging of uordetected arrays renders image data amenable to extraction by highly developed digital image processing proce dures Grids specifying target locations can be readily overlaid on the images Local sampling of background can be used to specify a threshold which true signal must exceed Mathematical morphology methods can be used to predict the likely shape and I u 39 M Fig 3 Atomic force microscopy of DNA on a microarray This is a micrograph of a portion of a hybridization probe from a yeast microarray taken after the array was subjected to hybridization The DNA is clearly deposited at a suffi cient density to allow many kinds of strand to strand interactions The width of the picture represents a scanned distance of 2 pm Image kindly provided by J DeRisi Stanford and E Carr Hewlett Packard nature genetics supplement 0 volume 21 vjanuary1999 x 1999 Nature America Inc httpIgeneticsnaturecom 1999 Nature America Inc httpgeneticsnaturecom review Fig 4 Detection schemes and applications of cDNA 1 microarrays Quantitative changes in gene expres l0 sion can be detected using several schemes for 1000 which the limits of detection vary a Direct incor poration of fluorescent nucleotides into the cDNA 00 target can be used to examine expression profiles 10 luorescence direc inoorpom ion from 10 ug or more of total RNA Indirect fluores a 1 lllore c39lce quotld39rfe39cl 0m 2 cence as well as target and signal amplification and 3 3399 Signal Ampm39ca39on radioactivity on the other hand can be used to 2 0i quot030 my 000i 2 detect expression profilesfrom as little as 50 ng of E 001 104 total RNA This detection limit allows for the in g E vestigation of expression profiles from numerous 39 l00l detection lUs biological sources including cell culture clinical 104 We biopsies including autopsy material and histologi cal samples b Improvements in technology will permit the detection of expression profiles from less than 50 ng of total RNA increasing the utility of the technology with respect to studies in devel opment The limits of the various techniques are constantly changing and this chart is meant only to illustrate of current performance levels 105 no of oeiis 1 mg of tissue 107 placement of the hybridization signal By applying these methods it is possible to accurately detect even weak signals27 and extract a mean intensity above background for the target In contrast extraction of data from lm or phosphoreimage representations of radioactive hybridizations presents many dif culties for image analysis If the array is on a membrane there is frequently non linear warping of the matrix which means that the observed array will not have the strict geometric regularity of an array printed to a stiff matrix such as glass This introduces dif culty in developing highly accurate grids to specify target locations The spread of detectable particles from a disintegrating nuclide to the detector is highly sensitive to variations in distance between source and detector and produces a smooth transition from the highest levels of intensity to background This ensures that the image produced by radioactive exposure is composed of sections at many focal planes and renders impossible the appli cation of single simple pointespread functions to reconstitute a focused representation of the data The smoothness of the tran sition from maximum signal intensity to background signal intensity makes consideration of local background for each sig nal a dif cult proposition as one does not observe an abrupt readily discerned transition between signal and background but a smooth curve without a sharp derivative In carrying out comparisons of expression data using measure ments from a single array or multiple arrays the question of nore malizing data arises All experiments are carried out under conditions of a large excess of immobilized probe relative to labelled target The kinetics of hybridization are therefore pseudoe rst order and intereprobe competition is not a factor Under these conditions the linear differences arising from exact amount of applied target extent of target labelling ef ciencies of uor excitation and emission and detector ef ciency can be come pounded into a single variable and the information from each detection channel normalized It is best to achieve normalization by adjusting the sensitivity of detection photomultiplier voltage with uorescence or exposure time with radioactivity so that the measurements occupy the same dynamic range in the detector There are essentially two strategies that can be followed in carrye ing out the normalization One is based on a consideration of all of the genes in the sample and the other on a designated subset expected to be unchanging over most circumstances In either case variance of the normalizing set can be used to generate estie nature genetics supplement 0 volume 21 o january 1999 limit 107 10 100 1000 104 105 105 107 106 109 10 s 10 5 10 0001 001 01 l 10 100 amoun ol 5 ar ing ma erial developmen al 5 udies his ological samples clinical bio sies lt E gt cellculure 4 gt mates of expected variance leading to predicted con dence inter vals In instances of closely related samples the transcript level of many genes will remain unchanged making global normalization a useful tool As samples become more divergent the fraction of genes showing altered transcript levels increases and global nore malization yields a poorer estimate of normalization than would be achieved using a subset of constantly expressed genes Explicit methods have been developed which make use of a subset of genes for normalization and extract from the variance of this subset statistics for evaluating the signi cance of observed changes in the complete dataset27 An aspect common to all array techniques is the extent of reliability and variance in measurements So far most array methods have been validated by probing northern blots of the biological samples As with sequencing the best comparisons and measures of reliability can be made only when large data sets containing significant repetitions and overlapping data are freely available One can however clearly envisage strengths and weaknesses The simple and highly determined nature of immobilized hybridization probes in oligonucleotide arrays make them likely to yield the highest level of reproducibility of absolute measurement for a given element The ability of cDNA arrays to achieve elementebyeelement normalization with two colour uorescence detection and to use a single highly speci c immobilized probe could provide the most accurate measure ments of relative expression levels All methods should readily disclose large changes in transcript levels among those genes readily detected Data management and mining All array methods require the construction of databases for the management of information on the genes represented on the array the primary results of hybridization and the construction of algorithms to make it possible to examine the outputs from single and multiple array experiments ref 27 see also page 51 of this issue ref 28 Methods x 1999 Nature America Inc httpIgeneticsnaturecom review 1999 Nature America Inc httpgeneticsnaturecom vide useful insights into the molecular pathogenesis of a variety of diseases ref 29 see also page 48 ofthis issue ref 30 It will not however deliver the kind of intimate understanding of the highly intererelated control circuitry that is necessary to achieve true understanding of genome function A number of recent publica tions suggest that to achieve this objective we should reconsider our perception of transcriptional control as a simple oneoff switch to a model whereby control is analogous to a highly gated logic circuit where numerous often contradictory inputs are summed to produce a response31 33 To reach these goals bioloe 1 Schuler GD et al A gene map of the human genome Science 214 54e546 1996 Henikoff s etal Genefamili Science 278 609614 1997 Schena M shalon D Davis RW amp Brown PC Quantitative monitoring of gene expression patterns with a complementary DNA microarray Science 210 467470 1995 DeRisi J et al Use of a cDNA microarray to analyse gene expression patterns in human cancer Nature Genet 14 457460 1996 DeRisi JL lyer VR amp Brown Po Exploring the metabolic and genetic control of gene expression on a genomic scale Science 278 68e686 1997 Alwine JC Kemp DJ amp Stark GR Method for detection of specific RNAs in agarose gels by transfer to diazobenzyloxymethylpaper and hybridization with DNA probes Proc NatlAcad Sci USA 14 535M354 1977 Berk AJ amp sharp PA Sizing and mapping of early adenovirus mRNAs by gel electrophoresis of s1 endonucleasedigested hybrids Cell 12 721432 1977 Liang P amp Pardee AB Differential display of eukaryotic messenger RNA by means ofthe polymerase chain reaction Science 257 967e971 1992 9 Adams MD et al Complementary DNA sequencing expressed sequence tags and human genome project Science 252 16514656 1991 10 okubo K et al Large scale cDNA sequencing for analysis of quantitative and qualitative aspects of gene expression Nature Genet 2 173479 1992 11 Velculescu VE zhang L Vogelstein B amp Kinzler KW Serial analysis of gene expression Science 210 481L487 1995 12 Schena M etal Parallel human genome analysis microarray based expression monitoring of 1 000 genes Proc NatlAcad Sci USA 93 1061440619 1996 13 Lockhart DJ et al Expression monitoring by hybridization to high density oligonucleotide arrays Nature Biotechnol 14 16754680 1996 14 Lipshutz RJ lodor SPA Gingeras TR amp Lockhart DJ High density synthetic oligonucleotide arrays Nature Genet 21 20e24 1999 15 Southern E Mir K amp shchepinov M Molecular interactions on microarrays Nature Genet 21 5e9 1999 16 Heller RA et al Discovery and analysis of inflammatory disease related genes using cDNA microarrays Proc NatlAcad Sci USA 94 215M155 1997 17 Wodicka L Dong H Mittmann M Ho MH at Lockhart DJ Genome wide expression monitoring in Saccharomyces cerevisiae Nature Biotechnol 15 N the taxonomy of protein paralogs and chimeras w P m 9 e 9 14 gists must expand the arsenal of tools they use to analyse expres sion datairecruiting statisticians and mathematicians to con sider multivariant problems of a size never before attempted Acknowledgements A hast uf talented investiga tars have can tribu ted ta the NIH Mitruamzy Prujett including YIiang A Glatfelter G Guaden Kuhn M Buguski G Schiller O Emulaeva E Daugherty T Pahida P Smith S Leightun I Hudsun A Furnace S Amundsun S Zeithner C Xiang R Simun I DeRisi P Brown 13591367 1997 18 de Saizieu A etal Bacterial transcript imaging by hybridization of total RNA to oligonucleotide arrays Nature Biotechnol 16 4548 1998 19 Benson DA Boguski MS Lipman DJ amp Ostell J GenBank NucleicAcids Res 251e61997 20 Boguski Ms Lowe TM amp Tolstoshev CM dbESTalatabase for quotexpressed sequence tagsquot Nature Genet 4 332e333 1993 21 Bowtell DL Options availablFfrom start to finishifor obtaining expression data by microarray Nature Genet 21 25e32 1999 22 cheung vG et al Making and reading microarrays Nature Genet 21 1549 1999 23 Phillips J amp Eberwine JH Antisense RNA amplificatio linear amplification method for analyzing the mRNA population from single living cells Methods 10 283e288 1996 24 Chen JJ etal Profiling expression patterns and isolating differentially expressed genes by cDNA microarray system with colorimetry detection Genomics 51 313e3241998 25 Thorp HH Cutting out the middleman DNA biosensors based on electrochemical oxidation Trends Biotechnol 16 117421 1998 26 Marshall A amp Hodgson J DNAchips an array of possibilities Nature Biotechnol 16 27e31 1998 27 Chen Y Dougherty ER amp Bittner ML Ratio based decisions and the quantitative analysis of cDNA microarray images I Blamed Optics 2 361L374 1997 28 Ermolaeva o etal Data management and analysis for gene expression arrays Nature Genet 20 1923 1998 29 Khan J et al Gene expression profiling of alveolar rhabdomyosarcoma with cDNA microarrays Cancer Res 58 50095013 1998 30 Debouck C amp Goodfellow P DNA microarrays in drug discovery and development Nature Genet 21 4850 1999 McAdams HH amp Shapiro L Circuit simulation of genetic networks Science 269 65e656 1995 32 Yuh CH Bolouri H 8 Davidson EH Genomic cisregulatory logic experimental and computational analysis of a sea urchin gene Science 279 1 8964 902 1998 33 Evan G amp Littlewood T A matter of life and cell death Science 281 13174322 1998 3 nature genetics supplement 0 volume 21 o januavy 1999
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'