CoChat

The impact of quantitative optimization of hybridization conditions on gene expression analysis

Open access

The DOI matches its Crossref record; 52 of the 52 references it cites were found in a registry or an index.

Abstract

BACKGROUND: With the growing availability of entire genome sequences, an increasing number of scientists can exploit oligonucleotide microarrays for genome-scale expression studies. While probe-design is a major research area, relatively little work has been reported on the optimization of microarray protocols. RESULTS: As shown in this study, suboptimal conditions can have considerable impact on biologically relevant observations. For example, deviation from the optimal temperature by one degree Celsius lead to a loss of up to 44% of differentially expressed genes identified. While genes from thousands of Gene Ontology categories were affected, transcription factors and other low-copy-number regulators were disproportionately lost. Calibrated protocols are thus required in order to take full advantage of the large dynamic range of microarrays.For an objective optimization of protocols we introduce an approach that maximizes the amount of information obtained per experiment. A comparison of two typical samples is sufficient for this calibration. We can ensure, however, that optimization results are independent of the samples and the specific measures used for calibration. Both simulations and spike-in experiments confirmed an unbiased determination of generally optimal experimental conditions. CONCLUSIONS: Well calibrated hybridization conditions are thus easily achieved and necessary for the efficient detection of differential expression. They are essential for the sensitive pro filing of low-copy-number molecules. This is particularly critical for studies of transcription factor expression, or the inference and study of regulatory networks.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/1471-2105-12-73

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Simulation study of cross-hybridization effects. To examine the impact of cross-hybridization, we considered 100 genes each with zero, low, medium, or high differential expression (Z/L/M/H). The effects of cross-hybridization were tested by adding a small off set to half of the genes in the channel with lower expression values. The likelihood, Eq. (5), is shown relative to the non-contaminated case ('no xhyb') as a baseline. From a selective perturbation of individual gene groups ('xhyb Z', 'xhyb L', 'xhyb M', 'xhyb H') we could see that cross-hybridization to probes for non-expressed genes can inflate the objective function. On the other hand, the corresponding loss for genes with medium and high expression levels is considerably larger. The overall effect of cross-hybridization for biological samples which do not show a strong bias towards either down- or up-regulation is labelled 'xhyb unbiased'. The obtained degraded likelihood shows that, as required, cross-hybridization in general has a deleterious outcome as long as unsuitably biased biological samples are avoided.
Figure 1 Simulation study of cross-hybridization effects. To examine the impact of cross-hybridization, we considered 100 genes each with zero, low, medium, or high differential expression (Z/L/M/H). The effects of cross-hybridization were tested by adding a small off set to half of the genes in the channel with lower expression values. The likelihood, Eq. (5), is shown relative to the non-contaminated case ('no xhyb') as a baseline. From a selective perturbation of individual gene groups ('xhyb Z', 'xhyb L', 'xhyb M', 'xhyb H') we could see that cross-hybridization to probes for non-expressed genes can inflate the objective function. On the other hand, the corresponding loss for genes with medium and high expression levels is considerably larger. The overall effect of cross-hybridization for biological samples which do not show a strong bias towards either down- or up-regulation is labelled 'xhyb unbiased'. The obtained degraded likelihood shows that, as required, cross-hybridization in general has a deleterious outcome as long as unsuitably biased biological samples are avoided.CC BY · PubMed Central
Figure 2 Simulation of different biological samples. We consider biological samples with very different characteristics in order to confirm that the protocol assessment is independent of the particular samples used for calibration. The pie charts in the second column of this figure show the varying distributions of differential expression examined. They were obtained by randomly selecting different proportions of genes with zero, low, medium, or high differential expression (Z/L/M/H) from the original measurements. The performance of the competing protocols (A... E) is quantitatively assessed and ranked for each simulation, represented by the rows in the second column. The distribution of ranks (third column) then indicates to what degree the obtained ranking is independent of the samples used for calibration. A stable ranking means that the found optimum will equally apply to arbitrary future measurements. In the example illustrated here, the optimum 'C' was robustly ranked first, whereas the protocols 'A' and 'B' as well as 'D' and 'E' show similar average performances. The optimum 'C' was therefore reliably identified.
Figure 2 Simulation of different biological samples. We consider biological samples with very different characteristics in order to confirm that the protocol assessment is independent of the particular samples used for calibration. The pie charts in the second column of this figure show the varying distributions of differential expression examined. They were obtained by randomly selecting different proportions of genes with zero, low, medium, or high differential expression (Z/L/M/H) from the original measurements. The performance of the competing protocols (A... E) is quantitatively assessed and ranked for each simulation, represented by the rows in the second column. The distribution of ranks (third column) then indicates to what degree the obtained ranking is independent of the samples used for calibration. A stable ranking means that the found optimum will equally apply to arbitrary future measurements. In the example illustrated here, the optimum 'C' was robustly ranked first, whereas the protocols 'A' and 'B' as well as 'D' and 'E' show similar average performances. The optimum 'C' was therefore reliably identified.CC BY · PubMed Central
Figure 3 Illustration of N-fold cross-testing. Illustration of N-fold cross-testing based on an A vs B assay of six replicates with dye-swap. Slides One to Three measure A vs B, whereas slides Four to Six assess B vs A in dye swap. With six-fold cross-testing every slide is once used as the independent test sample. ROC curves, generalization accuracies, and bit rates are estimated on the test samples after inferring the model from the training set.
Figure 3 Illustration of N-fold cross-testing. Illustration of N-fold cross-testing based on an A vs B assay of six replicates with dye-swap. Slides One to Three measure A vs B, whereas slides Four to Six assess B vs A in dye swap. With six-fold cross-testing every slide is once used as the independent test sample. ROC curves, generalization accuracies, and bit rates are estimated on the test samples after inferring the model from the training set.CC BY · PubMed Central
Figure 4 ROC curves for different hybridization temperatures. The information content of the measurements is reflected by ROC curves for the prediction of sample type (male vs female). Prediction performance was obtained independently for each gene by 6-fold cross-testing (Fig. 3). The ROC curves show averages over all genes. Comparing performances at different hybridization temperatures, the largest areas under the curve were observed for 51°C and 50°C (b). More subtle differences of the ROC curves can be examined in the zoomed detail plot (right panel), which identifies 51°C as optimal hybridization temperature.
Figure 4 ROC curves for different hybridization temperatures. The information content of the measurements is reflected by ROC curves for the prediction of sample type (male vs female). Prediction performance was obtained independently for each gene by 6-fold cross-testing (Fig. 3). The ROC curves show averages over all genes. Comparing performances at different hybridization temperatures, the largest areas under the curve were observed for 51°C and 50°C (b). More subtle differences of the ROC curves can be examined in the zoomed detail plot (right panel), which identifies 51°C as optimal hybridization temperature.CC BY · PubMed Central
Figure 5 Rank distributions of different hybridization protocols. Pie charts illustrate the rank distributions of competing protocols from performance comparisons for 100 gene subsets that correspond to samples of different properties. Every subset was used for evaluating all eight protocols to obtain a ranking (see schema in Fig. 2). While some random variation between the observed protocol rankings must be expected, in 90% of all tests the hybridization temperature of 51°C performed best, with the remaining 10% split between the two 50°C hybridizations ('50a', '50b'). This verified the reliability of the original calibration. The remaining ranking was also very stable, with a larger uncertainty, as expected, only seen for the two hybridizations at 50°C. The high similarity of the two 50°C measurements further confirmed the reproducibility of the calibration process. In summary, the obtained ranking and, in particular, the obtained calibration optimum were robust and independent of the biological samples used for calibration.
Figure 5 Rank distributions of different hybridization protocols. Pie charts illustrate the rank distributions of competing protocols from performance comparisons for 100 gene subsets that correspond to samples of different properties. Every subset was used for evaluating all eight protocols to obtain a ranking (see schema in Fig. 2). While some random variation between the observed protocol rankings must be expected, in 90% of all tests the hybridization temperature of 51°C performed best, with the remaining 10% split between the two 50°C hybridizations ('50a', '50b'). This verified the reliability of the original calibration. The remaining ranking was also very stable, with a larger uncertainty, as expected, only seen for the two hybridizations at 50°C. The high similarity of the two 50°C measurements further confirmed the reproducibility of the calibration process. In summary, the obtained ranking and, in particular, the obtained calibration optimum were robust and independent of the biological samples used for calibration.CC BY · PubMed Central
Figure 6 Box plots of log fold changes. We investigate the distribution of the log fold change of genes identified as significantly up- regulated at the optimal hybridization temperature but lost when hybridizing one degree Celsius too high. The left box plot shows that genes with sufficiently large fold change could also be detected in hybridizations at 52°C. The right box plot contrasts this with the distribution for genes only identified as up-regulated at the optimal hybridization temperature. Genes with more subtle fold changes were therefore lost at the suboptimal hybridization temperature. This particularly affects small copy number molecules, including many regulators.
Figure 6 Box plots of log fold changes. We investigate the distribution of the log fold change of genes identified as significantly up- regulated at the optimal hybridization temperature but lost when hybridizing one degree Celsius too high. The left box plot shows that genes with sufficiently large fold change could also be detected in hybridizations at 52°C. The right box plot contrasts this with the distribution for genes only identified as up-regulated at the optimal hybridization temperature. Genes with more subtle fold changes were therefore lost at the suboptimal hybridization temperature. This particularly affects small copy number molecules, including many regulators.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 52 references

Research works printed but not linked: 2

  1. NCBI Genome Assemblies and Resources[ http://www.ncbi.nlm.nih.gov/genome/ ]
  2. Huber W, von Heydebreck A, Sultmann H, Poustka A, Vingron M: Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002, 18(Suppl 1):S96–104. · DOI · Paper page

Cited by

See all 17

Metadata

Authors
Sykacek, P., Kreil, D. P., Meadows, L. A., Auburn, R. P., Fischer, B., Russell, S., & Micklem, G.
Affiliations
  • BOKU University (AT)
  • University of Cambridge (GB)
Corresponding author
Sykacek, Peter
Venue
BMC Bioinformatics
Published
2011-03-14
Volume
12
Issue
1
Pages
73-73
ISSN
1471-2105 (ISSN-L)
OpenAlex
W2107210684
Cited by
13
References
52
Citations by year
2022: 1 · 2019: 3 · 2018: 2 · 2015: 1 · 2014: 1 · 2013: 3 · 2012: 2
FWCI
0.78
Type
journal article
Language
en
Open access
Licence
  • cc-by
  • creativecommons.org/licenses/by/2.0 (tdm)
Locations
  • BMC Bioinformatics (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • BOK:ePub (Universitätsbibliothek der Universität f Bodenkultur Wien) (submittedVersion)
  • Apollo (University of Cambridge) (submittedVersion, cc-by)
Funders
  • Wellcome Trust (GR067205MA)
  • Vienna Science and Technology Fund
  • Universität Wien
  • Isaac Newton Trust
  • Directorate for Biological Sciences
  • Medical Research Council (G81/555)
  • Biotechnology and Biological Sciences Research Council (IGF12434, 8/EGH16 106, G18877)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Molecular Biology Techniques and Applications
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
Keywords
oligonucleotide microarrays, hybridization temperature, differential gene expression, transcription factors, information maximization, spike-in experiments, regulatory network inference
MeSH terms
Animals, Calibration, Drosophila melanogaster, Female, Male, Nucleic Acid Hybridization, Software, Temperature, Likelihood Functions, Linear Models, Oligonucleotide Array Sequence Analysis, Gene Expression Profiling

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: The impact of quantitative optimization of hybridization conditions on gene expression analysis. https://app.cochat.ai/papers/10.1186/1471-2105-12-73
The paper
Sykacek, P., Kreil, D. P., Meadows, L. A., Auburn, R. P., Fischer, B., Russell, S., & Micklem, G. (2011). The impact of quantitative optimization of hybridization conditions on gene expression analysis. BMC Bioinformatics. https://doi.org/10.1186/1471-2105-12-73

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.