Lara Lusa1,2, Vera Cappelletti1, Manuela Gariboldi1,2, Cristina Ferrario1,2, Loris De Cecco1,2, James F. Reid1,2, Sara Toffanin1, Giuseppe Gallus3, Lisa M. McShane4, Maria Grazia Daidone1, Marco A. Pierotti1,2
1Department of Experimental Oncology,
Istituto Nazionale per lo Studio e la Cura dei Tumori, Milan
2Molecular Genetics of Cancer Group,
Fondazione Istituto FIRC di Oncologia Molecolare (IFOM), Milan
3Medical Statistics and Biometry,
Istituto Nazionale per lo Studio e la Cura dei Tumori, Milan - Italy
4Biometric Research Branch, National
Cancer Institute, Bethesda, MD - USA
Tóm tắt
We describe a microarray experiment using the MCF-7 breast cancer cell line in two different experimental conditions for which the same number of independent pools as the number of individual samples was hybridized on Affymetrix GeneChips. Unexpectedly, when using individual samples, the number of probe sets found to be differentially expressed between treated and untreated cells was about three times greater than that found using pools. These findings indicate that pooling samples in microarray experiments where the biological variability is expected to be small might not be helpful and could even decrease one's ability to identify differentially expressed genes.
