Using data from 243 breast cancer samples analyzed on two different microarray platforms, our in-house 29K array platform and Agilent SurePrint G3 microarray platform, we demonstrated the importance of detecting and tackling batch-related biases within datasets prior to data analysis.
Results from such analysis are consistent with clinical pathology IHC markers for ER, PR and HER2 and comparable to results from fresh/frozen specimens on traditional microarray platforms as demonstrated on independent published microarray-generated data sets (see below).
Amongst the microarray platforms, Agilent demonstrated the largest signal range and signal-to-background ratio and Illumina had the least values of these metrics, as seen in Fig. 2a-b.
This prognostic gene-expression signature was derived from two patient cohorts, and validated on four fully independent patient cohorts treated in North America, Europe and Australia, and profiled using different microarray platforms, thus demonstrating its robustness across clinical spectrum and assay platforms.
The robustness and reliability of the mouse NIA-15K cDNA microarray platform has been demonstrated in our previous work [ 11, 12] and by others [ 13- 15].
Overall, replication of the microarray results was high using the nCounter platform, demonstrating the reliability of each platform.
By validating the comparability between data from distinct microarray platforms we have demonstrated a tool that can enhance the statistical power provided by large sample sets.
The method can be extended to other 3' expression microarray platforms as we demonstrate with human data.
These analyses demonstrate that (1) BM-MSCs from AML patients and healthy donors showed no significant difference in mRNA expression as measured by Illumina's BeadChip microarray platform and (2) AML blasts demonstrate increased inter-patient heterogeneity and have different mRNA expression profiles as compared to their corresponding BM-MSCs and normal hematopoietic MNCs.
A subset of sera that was tested on the microarray platform was selected at random and demonstrated seroreactivity on the immunostrip platform.
