Strand Specificity of Arrays and RNAseq

Question

[ad]

Hi, I have some questions about the strand specificity of certain expression arrays and TCGA RNAseq data.

Does anybody know if any of the below are strand specific or not? This information doesn't seem to be easily accessible on the manufacturers website.

TCGA RNAseq, TCGA RNAseqV2 gene and exon expression data

Affymetrix HG-U133_Plus_2, HG-U133

Illumina HumanHT-12 V3.0 expression beadchip

Answer 1

[seidel]

The Affymetrix and Illumina technologies are strand-specific. They are both based on oligo-synthesis technologies - thus the strand required for detection is the strand that is produced. The TCGA data collection contains data from such a wide array of samples over time I doubt it if all the data sets are derived from strand-specific sequencing protocols, but I bet it's becoming more common, if not required for contemporary data sets. (There are methods for determining if a data set is strand specific).

[Chris Evelo]

Especially for older microarrays @seidel's answer is correct in principle but in practice often translates to "should be strand-specific". Microarray reporters were often found to be less well targeted than intended; hitting the wrong gene, hitting multiple genes and indeed sometimes targeting the wrong strand. Those were in fact among the reason for the custom microarray annotations that are available from Brainarray.