|
Supplementary Materials and Methods
To download this as a Word file (28kb), click here
Data Analysis and Retrieval. For microarray Method 1, data for each
experiment were extracted into Microsoft Excel after filtering out array
elements that yielded weak signal intensities: the sum of median intensities
for the two channels was required to be greater than 150, and the regression
correlation value (r2) for the two channels was required to be greater or
equal to 0.6.
For Method 2, data were stored and extracted from the Stanford
Microarray Database (1). Similar to Method I, low intensity signals were
removed from the data set by filtering those elements for which the ratio of
the signal to the background was greater than 1.5 for the Cy3 channel (total
RNA control), and greater than 1.0 for the Cy5 channel (pull-down RNA). In
addition, the regression correlation value (r2) for the two channels was
required to be greater or equal to 0.6. Furthermore, array elements were
removed from the data set if more than five of the ten experiments in this
method did not satisfy the requirements described above.
Since actual ratio values for enriched RNAs in individual pull-downs
were not comparable due to differences in pull-down methodology,
amplification and immunoprecipitation efficiencies, we used a method
identical to Lieb at al (2) whereby the median percentile rank for each array
element was calculated using ranked relative enrichment values for each
experiment.
Data Interpretation. Two of the 5 experiments from Method 1 were performed with microarrays containing coding and intergenic sequences, whereas the remainder were performed with microarrays containing only yeast coding sequences. To compare the experiments, all intergenic loci were eliminated before percentile ranks were assigned. Median ranks from all 5 experiments were then compared directly to determine median rankings.
The experiments performed with intergenic arrays suggested that several RNAs from noncoding regions potentially interact with the She proteins. However, only two experiments were performed with these types of arrays, and we are thus hesitant to attribute significance to these results at this stage. Some of these intergenic regions are adjacent to localized coding sequences (ASH1, IST2), while others are derived from overlapping sequencesÉ (as the arrays are derived from PCR products, a positive hybridization signal does not distinguish strands). Several, however, are unrelated to tested transcripts.
In addition to intergenic sequences, several transposable elements were identified as positives by our microarray screens. Because of the large number of candidates to investigate, we chose to focus our efforts on RNA derived from coding sequences. The relevance of She-interactions with intergenic and transposon-derived RNA remains to be determined.
1. Gollub J, Ball CA, Binkley G, Demeter J, Finkelstein DB, Hebert JM, Hernandez-Boussard T, Jin H, Kaloper M, Matese JC, Schroeder M, Brown PO, Botstein D, Sherlock G. (2003) The Stanford Microarray Database: data access and qualityassessment tools. Nucleic Acids Res 31(1):94-6
2. Lieb, J. D., Liu, X., Botstein, D. & Brown, P. O. (2001) Nat Genet 28, 327-34.
|