Limits...
Correlating measurements across samples improves accuracy of large-scale expression profile experiments.

Alvarez MJ, Sumazin P, Rajbhandari P, Califano A - Genome Biol. (2009)

Bottom Line: Gene expression profiling technologies suffer from poor reproducibility across replicate experiments.However, when analyzing large datasets, probe-level expression profile correlation can help identify flawed probes and lead to the construction of truer probe sets with improved reproducibility.We test and validate our approach on Affymetrix microarrays and outline their future adaptation to other technologies.

View Article: PubMed Central - HTML - PubMed

Affiliation: Joint Centers for Systems Biology, Columbia University, 2960 Broadway, New York, NY 10027-6900, USA. malvarez@c2b2.columbia.edu

ABSTRACT
Gene expression profiling technologies suffer from poor reproducibility across replicate experiments. However, when analyzing large datasets, probe-level expression profile correlation can help identify flawed probes and lead to the construction of truer probe sets with improved reproducibility. We describe methods to eliminate uninformative and flawed probes, account for dependence between probes, and address variability due to transcript-isoform mixtures. We test and validate our approach on Affymetrix microarrays and outline their future adaptation to other technologies.

Show MeSH

Related in: MedlinePlus

Cleaner probe clusters for six genes profiled on an exon array. Examples of putative Cleaner mRNA isoforms identified by using 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays. The plots show the hybridization position for each probe on known and predicted mRNA isoforms for each gene.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2812950&req=5

Figure 6: Cleaner probe clusters for six genes profiled on an exon array. Examples of putative Cleaner mRNA isoforms identified by using 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays. The plots show the hybridization position for each probe on known and predicted mRNA isoforms for each gene.

Mentions: Cleaner is directly applicable for analyzing Affymetrix exon arrays and is expected to produce alternative clusters for a larger set of genes. Figure 6 illustrates six examples of Cleaner probe clusters associated with known transcript isoforms (FAM13A, ELMO1, TPD52L2, INO80C) and Cleaner predicted isoforms (GATA6 and MAPK8IP1) using expression data from 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays [29]. Note that alternative cluster probe positions for TPD52L2 and INO80C are interleaved. To emphasize that expression estimates for the alternative probe clusters are indeed very different, we compare expression estimates using the alternative clusters for the six genes in Figure S6 in Additional file 1.


Correlating measurements across samples improves accuracy of large-scale expression profile experiments.

Alvarez MJ, Sumazin P, Rajbhandari P, Califano A - Genome Biol. (2009)

Cleaner probe clusters for six genes profiled on an exon array. Examples of putative Cleaner mRNA isoforms identified by using 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays. The plots show the hybridization position for each probe on known and predicted mRNA isoforms for each gene.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2812950&req=5

Figure 6: Cleaner probe clusters for six genes profiled on an exon array. Examples of putative Cleaner mRNA isoforms identified by using 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays. The plots show the hybridization position for each probe on known and predicted mRNA isoforms for each gene.
Mentions: Cleaner is directly applicable for analyzing Affymetrix exon arrays and is expected to produce alternative clusters for a larger set of genes. Figure 6 illustrates six examples of Cleaner probe clusters associated with known transcript isoforms (FAM13A, ELMO1, TPD52L2, INO80C) and Cleaner predicted isoforms (GATA6 and MAPK8IP1) using expression data from 55 human glial brain tumor samples hybridized on huex10stv1 Affymetrix exon arrays [29]. Note that alternative cluster probe positions for TPD52L2 and INO80C are interleaved. To emphasize that expression estimates for the alternative probe clusters are indeed very different, we compare expression estimates using the alternative clusters for the six genes in Figure S6 in Additional file 1.

Bottom Line: Gene expression profiling technologies suffer from poor reproducibility across replicate experiments.However, when analyzing large datasets, probe-level expression profile correlation can help identify flawed probes and lead to the construction of truer probe sets with improved reproducibility.We test and validate our approach on Affymetrix microarrays and outline their future adaptation to other technologies.

View Article: PubMed Central - HTML - PubMed

Affiliation: Joint Centers for Systems Biology, Columbia University, 2960 Broadway, New York, NY 10027-6900, USA. malvarez@c2b2.columbia.edu

ABSTRACT
Gene expression profiling technologies suffer from poor reproducibility across replicate experiments. However, when analyzing large datasets, probe-level expression profile correlation can help identify flawed probes and lead to the construction of truer probe sets with improved reproducibility. We describe methods to eliminate uninformative and flawed probes, account for dependence between probes, and address variability due to transcript-isoform mixtures. We test and validate our approach on Affymetrix microarrays and outline their future adaptation to other technologies.

Show MeSH
Related in: MedlinePlus