Statistical evaluation of SAGE libraries: consequences for experimental design

Abstract

Since the introduction of serial analysis of gene expression (SAGE) as a method to quantitatively analyze the differential expression of genes, several statistical tests have been published for the pairwise comparison of SAGE libraries. Testing the difference between the number of specific tags found in two SAGE libraries is hampered by the fact that each SAGE library is only one measurement: the necessary information on biological variation or experimental precision is not available. In the currently available tests, a measure of this variance is obtained from simulation or based on the properties of the tag distribution. To help the user of SAGE to decide between these tests, five different pairwise tests have been compared by determining the critical values, that is, the lowest number of tags that, given an observed number of tags in one library, needs to be found in the other library to result in a significant P value. The five tests included in this comparison are SAGE300, the tests described by Madden et al. ( Oncogene 15: 1079–1085, 1997) and by Audic and Claverie ( Genome Res 7: 986–995, 1997), Fisher’s Exact test, and the Z test, which is equivalent to the chi-squared test. The comparison showed that, for SAGE libraries of equal as well as different size, SAGE300, Fisher’s Exact test, Z test, and the Audic and Claverie test have critical values within 1.5% of each other. This indicates that these four tests will give essentially the same results when applied to SAGE libraries. The Madden test, which can only be used for libraries of similar size, is, with 25% higher critical values, more conservative, probably because the variance measure in its test statistic is not appropriate for hypothesis testing. The consequences for the choice of SAGE library sizes are discussed.

Transcriptional Profiling Reveals Novel Markers of Liver Fibrogenesis

“…SAGE data analysis was performed using USAGE V2 software, a World Wide Web-based application developed in the Bioinformatics Laboratory of our Institute (17), for extraction of single tags from sequence data and subsequent identification on the NCBI human gene data base. The three obtained nonnormalized SAGE libraries were evaluated using SAGEstat (18), and differences were evaluated as being statistically significant with p values less than 0.01. To further study tag identification and expression, the NCBI/CGAP/SAGEMAP program was used on the World Wide Web at www.ncbi.nlm.nih.gov/SAGE/ or cgap.nci.nih.gov/SAGE/.…”

Statistical Design and Analysis of RNA Sequencing Data

“…See Man et al (2000), Romualdi et al (2001), and Ruijter et al (2002) (Fisher 1935b), which fixes the marginal totals of the 2 3 2 table and tests differential expression using the hypotheses …”

“…This phenomenon as related to RNA-Seq data is discussed in detail in Bloom et al (2009). The methods used by Kal et al (1999) (a test of the equality of two binomial proportions) and Audic and Claverie (1997) (a Bayesian model with a Poisson likelihood and a flat prior on the mean) may also be used, although comparisons between these, and other, approaches to Fisher's exact test (Man et al 2000;Romualdi et al 2001;Ruijter et al 2002) show marginal differences in performance.…”

The Transcriptome Profile of Human Embryonic Stem Cells as Defined by SAGE

“…Since SAGE libraries comprise discrete data, they can be subjected to pairwise comparison to statistically analyze the differential expression of genes [25] and to generate a comparative digital gene expression profile [24].…”