By continuing to use our site, or clicking "Continue," you are agreeing to our Cookie Policy Continue. Figure 1. There are issues that complicate each step of microarray gene expression analysis. Figure 2. A, Tissue sections are processed and placed on a microscope slide under a thin, transparent thermoplastic film, which is attached to a movable cap.
Visualizing the tissue microscopically, a short-duration, focused pulse from an infrared carbon-dioxide laser is used to activate and melt the film to selectively adhere cells within targeted areas of interest. B, When the cap is lifted, the film, with selected cells still bound, is removed from the tissue section for further processing to retrieve cellular materials eg, DNA, RNA, proteins.
Figure 3. B, A dual-labeled fluorogenic probe with a higher melting point than the PCR primers used for extension is annealed to the target sequence between the forward and reverse PCR primers.
Fluorescence is measured continuously throughout the PCR amplification in real time and is proportional to the amount of PCR product generated in each cycle. The sequence of the human genome. Google Scholar. Fields S. The future is function. Nat Genet. Lander ES. The new genomics. Liang P, Pardee AB. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Serial analysis of gene expression. Expression monitoring by hybridization to high-density oligonucleotide arrays.
Nat Biotechnol. A new dimension for the human genome project. Cytochrome P 2E1 polymorphism in gastric cancer in Brazil. Cancer Epidemiol Biomarkers Prev. Isolation of rat pgp3 cDNA. Biochim Biophys Acta. Loss of heterozygosity in normal tissue adjacent to breast carcinomas.
Barrett's esophagus. Cancer Res. Allele-specific chromosome 3p deletions occur at an early stage in the pathogenesis of lung carcinoma. Br J Cancer. The genetics of cancer—a 3D model. Identification of differentially expressed genes in human prostate cancer using subtraction and microarray.
Microarray-based expression profiling in prostate tumors. In Vivo. Prostate tumours from an Asian population. Ki-ras point mutation and p53 expression in human pancreatic cancer. Pettaway CA. J Natl Med Assoc. Expression of the endostatin gene in epithelial ovarian cancer. Clin Cancer Res. Saida T. Recent advances in melanoma research. J Dermatol Sci. Hegde U, Wilson WH. Gene expression profiling of lymphomas. Curr Oncol Rep. Liu L, Yang K. A study on C-erbB2, nm23 and p53 expressions in epithelial ovarian cancer and their clinical significance [in Chinese].
Detection of differentially expressed genes in human colon carcinoma cells treated with a selective COX-2 inhibitor. Association between expression of the MRP3 gene and exposure to platinum drugs in lung cancer. Int J Cancer. Molecular classification of cancer. Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks.
Nat Med. Recursive partitioning for tumor classification with gene expression microarray data. Molecular classification of cutaneous malignant melanoma by gene expression profiling. Analysis of gene expression profiles in normal and neoplastic ovarian tissue samples identifies candidate molecular markers of epithelial ovarian cancer.
Systematic variation in gene expression patterns in human cancer cell lines. Altered patterns of gene expression in response to myocardial infarction. Circ Res.
Abdellatif M. Leading the way using microarray. Laser capture microdissection. In vivo gene expression profile analysis of human breast cancer progression. Effects of ethanol and paraformaldehyde on RNA yield and quality. Small-sample total RNA purification. The application of DNA microarrays in gene expression analysis. J Biotechnol. Burgess JK. Gene expression studies using microarrays.
Clin Exp Pharmacol Physiol. Alterations of gene expression during colorectal carcinogenesis revealed by cDNA microarrays after laser-capture microdissection of tumor tissues and normal epithelia.
The "chip" as a specific genetic tool. Biol Psychiatry. A genome-wide transcriptional analysis of the mitotic cell cycle. Mol Cell. Cellular clocks and oscillators. Int Rev Cytol. Physiol Genomics. Claverie JM. Computational methods for the identification of differential and coordinated gene expression. Hum Mol Genet. Comparison of normal human skin gene expression using cDNA microarrays. Wound Repair Regen.
Normalization strategies for cDNA microarrays. Nucleic Acids Res. Many methods for visualization, quality assessment, and data normalization have been developed see [9] for a review, Text S1 , and Figure S1. Clustering is a way of finding and visualizing patterns in the data. Many papers and indeed books have been written on this topic see e. Different methods highlight different patterns, so trying more than one method can be worthwhile. Note that while clustering finds predominant patterns in the data, those patterns may not correspond to the phenotypic distinction of interest in the experiment.
To identify gene expression patterns related to this distinction, more directed methods are appropriate. There are many commercial packages for microarray analyses, and we have by no means evaluated all of them. However, commercial tools can be expensive, and we find many that we have tried to have limited flexibility. Fortunately, in the past few years a number of Web-based tools and open-source software packages for microarray data analysis have become available see below and Text S1 , and we recommend taking advantage of them.
One common strategy is to create a custom data analysis pipeline using statistical analysis software packages such as Matlab or R. Both allow great flexibility, customized analysis, and access to many specialized packages designed for analyzing gene expression data. Not only is R freely available, but it also allows the use of BioConductor [14] , a collection of R tools including many powerful current gene expression analysis methods written and tested by experts from the growing microarray community.
The fundamental goal of most microarray experiments is to identify biological processes or pathways that consistently display differential expression between groups of samples. While the exact approach depends in part on the design of the experiment, there are two broad approaches to detecting differential expression. The first examines each gene or transcript individually to find genes that, by themselves, have statistically significant differences in expression between samples with different phenotypes or characteristics.
The set of genes thus identified is then examined for over-representation of specific functions or pathways [15]. A powerful alternative is to identify groups of functionally related genes ahead of time and to test whether these gene sets—as a group—show differential expression [16] — [18]. Both of these approaches can be effective, and sometimes the combination of the two is stronger than either alone [19]. One crucial issue for all microarray analysis methods is adjusting for multiple testing [20].
Each statistical test reports the probability of seeing the observed test score by chance under the null hypothesis that there is no difference in expression related to the phenotype being studied. A range of methods to adjust for multiple testing are available see [21] for an overview. Once a list of differentially expressed genes has been assembled, some functional analysis is essential for interpreting the results. There are many tools available to identify pathways or biological functions that are over-represented in a given gene list.
Again, adjustment for multiple testing may be desirable, although complex dependencies between pathways make finding an appropriate adjustment method controversial [23]. A good review of the earlier tools that discusses many of the statistical issues is [15]. An alternative to the individual-gene analysis workflow is to consider entire gene sets or pathways together when looking for differential expression. There are many approaches that do this e.
Gene set analysis can be advantageous because it can detect subtle changes in gene expression that individual gene analyses can miss, and because it combines identification of differential expression and functional interpretation into a single step. The disadvantage of this method is that appropriate gene sets need to be known ahead of time.
When studying a biological process that is still poorly understood, an individual gene method may be more appropriate, as it allows for the opportunity of implicating hitherto unexpected genes and gene sets. Given that gene set analysis is more sensitive and therefore potentially more powerful, a greater effort in defining the pathways needed to support this approach is warranted.
Toward this end, GSEA's gene set database incorporates some computationally derived gene sets, including expression neighbors of known cancer genes [17] and network modules mined from a large collection of expression data [27].
Related work has used conserved coexpression [28] or differential coexpression [29] to discover new functional modules. Much has also been written about sample classification using microarray data see review [13] but, with a few exceptions [30] , [31] , microarrays themselves have not been embraced as diagnostic tools. Rather, they have been used to identify smaller sets of predictive genes or pathways that might, when assessed by other technologies, aid in diagnosis or stratification of samples.
A huge range of machine learning methods [11] , [12] can be applied to the related classification problems. Most people intent on doing this write their own code but see Text S1 for an alternative.
We note that simpler classification tools often perform as well as, and generalize better than, more complex ones [32]. It has been our goal in this brief review to demonstrate that it is currently feasible for researchers with no previous experience to incorporate microarray analyses in their studies. The field is now reasonably mature, with available software and tools to make data analysis manageable by nonexperts. That said, newcomers to the field should be aware that the data analysis will require a dedicated commitment of time and effort that generally substantially exceeds that of data generation.
We strongly recommend that researchers do the work to familiarize themselves with the relevant analytical literature before beginning, or even designing, the experiment.
It has been speculated that microarray technology will soon be superseded by next-generation sequencing, in which the transcripts are directly sequenced by low-cost, high-throughput sequencing technologies [33]. However, currently, next-generation whole-transcriptome sequencing is still quite expensive and in its relative infancy. Its cost scales proportionally with its ability to assess low-abundance transcripts, as sufficient depth of sequencing must be performed. Do you want to LearnCast this session?
This article has been posted to your Facebook page via Scitable LearnCast. Change LearnCast Settings. Scitable Chat. Register Sign In.
For example, researchers believe that mutations in the genes BRCA1 and BRCA2 cause as many as 60 percent of all cases of hereditary breast and ovarian cancers.
But there is not one specific mutation responsible for all of these cases. Researchers have already discovered over different mutations in BRCA1 alone. The chip consists of a small glass plate encased in plastic. Some companies manufacture microarrays using methods similar to those used to make computer microchips.
On the surface, each chip contains thousands of short, synthetic, single-stranded DNA sequences, which together add up to the normal gene in question, and to variants mutations of that gene that have been found in the human population.
0コメント