Biology Reference
In-Depth Information
This methodology encompasses the techniques of SAGE [7,8] and
Massively Parallel Signature Sequencing (MPSS) [3,9]. MPSS repre-
sents a powerful alternative to microarray technologies. It provides a
sensitive measure of gene expression without requiring a priori knowl-
edge of transcribed sequences. The MPSS process is complex; from the
extraction of the total RNA to the quantification of transcripts, there are
a number of steps that contribute to the total noise. One significant
difference between MPSS and microarray technologies is that MPSS
produces data in digital format, whereas microarray data is analog [10].
(While the spot intensities obtained from microarray technologies are
digital, these values do not correspond to a specific number of target
molecules bound to a particular spot; that is, the data is a digital repre-
sentation of a signal that is analog relative to the intensity resolution of
the optical scanner.)
In this chapter, techniques for identifying and quantifying sources of
noise through the analysis of replicate experiments [6,11] are discussed.
In utilizing these techniques, one systematically studies the variations
between equivalent experiments which, in the absence of experimental
variability and noise, would yield identical results. An analysis is pre-
sented here of replicate experiments carried out with two systems:
Affymetrix GeneChip microarrays [12] and MPSS technology [13]. In
presenting these analyses, the intent is both to illustrate the general
technique and to observe the inherent differences between the analysis
of analog and digital gene expression data. Finally, it is shown how
noise analysis provides an effective method for quantifying the relative
sensitivity of different experimental techniques.
QUANTITATIVE NOISE ANALYSIS OF DNA MICROARRAY
EXPERIMENTS
The data and results presented in this section are largely derived from
the quantitative study of noise in Affymetrix GeneChips presented
in ref. [12], where more details on sample preparation and data pro-
cessing can be found. The data used in this section can be downloaded
Materials and Methods
In this section an analysis of the experimental noise introduced during
the sequential processing steps in high-density oligonucleotide-based
microarray (Affymetrix) assays is presented. Elucidating the sources of
noise may be of help in identifying those experimental steps that need
to be modified to improve the signal to noise ratio. The results pre-
sented here show that it is the hybridization (including the subsequent
readout) step, as opposed to the sample preparation step, where most
of the noise originates. Based on these results, a data analysis method
Search WWH ::
Custom Search