George I. Lambrou (University of Athens, Greece), Maria Braoudaki (University of Athens, Greece), Emmanouil Sifakis (Karolinska Institutet, Sweden) and Apostolos Zaravinos (Karolinska Institutet, Sweden)
Copyright: © 2015 |Pages: 14
DOI: 10.4018/978-1-4666-5888-2.ch552
OnDemand PDF Download:
No Current Special Offers

Chapter Preview



DNA microarrays detect patterns of gene expression, therefore they can be used for acquiring such “images” and the induction of conclusions on cell state (Diehn, Alizadeh, & Brown, 2000). cDNA microarrays have been used for a plethora of experiments: virtually any property of a DNA sequence which can be experimentally modified, may be determined as far as its differential expression is concerned, and this can be performed on thousands of sequences simultaneously. In that sense, microarrays have been used in gene mapping studies (Comparative Genomic Hybridization, CGH microarrays), mutational analysis and monitoring changes in gene expression of the genome (Eisen & Brown, 1999).

Research questions that can be answered with DNA microarrays are related mainly to the investigation of gene expression. They can compare the relative abundance of mRNA of a gene under investigation, between two different cells or tissue samples. For example, an experiment could compare cells before and after an experimental intervention, or at successive moments of a specific process, or between stages of differentiation or mRNA expressed in a mutant cell compared to that of wild type. This would be the simplest type of experiment.

Key Terms in this Chapter

Supervised Learning: Which uses a predisposed knowledge on the data set and based on this classifies the data.

Spotted Microarrays: The first arrays that have been created. These are produced by use of a robotic system known as “printer.”

Biological Replicates: Repetitions of different samples, but derived from the same question, on different arrays.

Class Discovery: A global gene expression profile is analyzed in order to discover subsets with common characteristics.

Class Prediction: Discovery of the class to which a sample belongs (based on its gene expression profile).

Loop-Design: Where all samples and references are compared with each other in all possible combinations.

Reporters: Nucleic acid sequences printed on a microarray slide.

Dye-Swap Design: Sample and reference are interchanged and hybridized with different dyes (e.g. in one experiment reference is labelled with Cy3 and sample with Cy5 and in a simultaneous experiment the same reference is labelled with Cy5 and the sample with Cy3).

In Situ Synthesis Microarrays: Reporters are synthesized at the very exact moment they are printed. Synthesis occurs through covalent bonding between the hydroxyl group of 5' hexose carbon and the phosphor group of the next nucleotide.

Reference Design: All samples are compared to one reference or a pool of references.

Technical Replicates: Repetition of the same experiment, with the same samples on a different array.

Unsupervised Learning: Includes the calculation of correlations among data without supervision or external help.

Class Comparison: Discovery of differences in expression levels between two or more samples.

Complete Chapter List

Search this Book: