The HuGeneFL and HG-U95A oligonucleotide microarrays from Affymetrix are two successive generations of the platform for human expression profiling. Because the latter microarray (one of a 5-chip set) is a superset of the former generation, one would expect that probe sets in one generation would have the same expression pattern as measured in the next generation of microarrays. However, as demonstrated in figure 7.6, there are a large number of probe sets that have very poor correlation when the same RNA sample is hybridized identically and in the same laboratory to the two generations of microarrays.
-0.2 0 0.2 0.4 Correlation coefficient (r)
Figure 7.6: Poor measurement reproducibility across microarray generations. Two generations of Affymetrix microarrays were used to the gene expression of seven samples. Each of the 8075 common probe sets thus had seven measurements from both the older HuGeneFL and newer HG-U95A microarrays. The line represents the distribution of correlation coefficients of each probe set. Though the exact correlation coefficient may not be useful, it is notable that over 25% of the probe sets showed measurements that were zero or negatively correlated across generations. (Derived from Nimgaonkar et al. .)
The most likely explanation for this disparity is that Affymetrix designed the HG-U95A array to use a set of probes or oligonucleotides for each probe set that may only partially overlap the set of probes used in the HuGeneFL arrays. Presumably this reflects improved oligonucleotide selections based on greater knowledge of the genome and accumulated empirical results of the specificity and sensitivity of particular probes. However, because Affymetrix has continued to maintain secrecy around the sequence of each oligonucleotide in a probe set, we cannot assess this directly. Affymetrix does release a table mapping the probes (within a probe set) in one generation to the next. This allows a comparison of how well probe sets correlate as a function of the number of shared probes (see figure 7.7). At a threshold of three or more probes in common per probe set, the correlation is substantially improved. This may explain in large part why the overall correlation between the two generations of microarrays is so poor.
Number of probe pairs shared between generations Figure 7.7: Measurement reproducibility of probe sets across microarray generations, by the number of probe pairs in common. A total of 8075 probe sets are deemed in common by Affymetrix between the older HuGeneFL and the newer HG-U95A microarrays. However, each probe set has a varying number of common probe pairs, from none to all 16. The bar graph represents the distribution of the 8075 probe sets, and the line represents the correlation coefficients of each set of probe sets across generations. For example, one finds a correlation coefficient of .77 when plotting all 6412 probe sets with one probe pair in common across microarray generations. (Derived from Nimgaonkar et al. .)
Was this article helpful?