Skip to main content

Table 3 Pearson's correlations between Affymetrix and Codelink data

From: Transcript-based redefinition of grouped oligonucleotide probe sets using AceView: High-resolution annotation for microarrays

Cross-platform Comparison

 

Pearson's correlations

 

N

D1/C1

D2/C2

D3/C3

UniGene-based match

7,399

0.55

0.55

0.58

AceView-based match

4,064

0.55

0.57

0.61

   Probe sets with size ≥ 4

3,694

0.59

0.61

0.63

   Probe sets with size < 4

370

0.34

0.37

0.39

RefSeq-based match*

7,801

0.60

0.61

0.64

  1. The D1-D3 and C1-C3 are replicates in the treatment group and the control group respectively. The correlation is calculated on log-ratios, where the pairing of technical replicates is arbitrary. The best concordance between two platforms was observed using AceView-based transcript-level probe sets (with size ≥ 4) and RefSeq based probe sets.
  2. *Using RefSeq-based mappings by Dai et al [13]