Limits...
Hybridization thermodynamics of NimbleGen microarrays.

Mueckstein U, Leparc GG, Posekany A, Hofacker I, Kreil DP - BMC Bioinformatics (2010)

Bottom Line: We demonstrate the benefits of an improved model for microarray hybridization and assess the relative contributions of the probe-target binding strength and the different competing structures.This analysis illustrates the importance of considering both the effects of probe-target binding strength and the different competing structures.Besides their relevance for the design of new arrays, our results demonstrate the value of improving thermodynamic models for the read-out and interpretation of microarray signals.

View Article: PubMed Central - HTML - PubMed

Affiliation: WWTF Chair of Bioinformatics, Boku University Vienna, Muthgasse 18, 1190 Vienna, Austria. Ulrike.Mueckstein@boku.ac.at

ABSTRACT

Background: While microarrays are the predominant method for gene expression profiling, probe signal variation is still an area of active research. Probe signal is sequence dependent and affected by probe-target binding strength and the competing formation of probe-probe dimers and secondary structures in probes and targets.

Results: We demonstrate the benefits of an improved model for microarray hybridization and assess the relative contributions of the probe-target binding strength and the different competing structures. Remarkably, specific and unspecific hybridization were apparently driven by different energetic contributions: For unspecific hybridization, the melting temperature Tm was the best predictor of signal variation. For specific hybridization, however, the effective interaction energy that fully considered competing structures was twice as powerful a predictor of probe signal variation. We show that this was largely due to the effects of secondary structures in the probe and target molecules. The predictive power of the strength of these intramolecular structures was already comparable to that of the melting temperature or the free energy of the probe-target duplex.

Conclusions: This analysis illustrates the importance of considering both the effects of probe-target binding strength and the different competing structures. For specific hybridization, the secondary structures of probe and target molecules turn out to be at least as important as the probe-target binding strength for an understanding of the observed microarray signal intensities. Besides their relevance for the design of new arrays, our results demonstrate the value of improving thermodynamic models for the read-out and interpretation of microarray signals.

Show MeSH
Importance ranking of thermodynamic probe properties including target-side modeling. Importance ranking of thermodynamic probe properties including target-side modelling. The left hand figure (A) shows the Guide ranking for the clearly expressed genes showing no cross-hybridization from Wei et al. [23], whereas the right one shows the results for the Pozhitkov et al. [26] dataset (B). ΔG stands for the effective interaction energy including relevant competing intra- and inter-molecular effects. ΔGp labels the free energy of the probe secondary structure, ΔGt the free energy of the probe binding site secondary structure in the target, ΔGh the free energy of the probe-target duplex, and ΔGpp the free energy of probe-probe dimerization within the same probe feature. 'Tm' stands for the melting temperature. Error bars are from bootstrap re-sampling of 90% of all probes. Below the x-axis, the Spearman rank correlation of predictions to the observed signal intensity is shown. All correlations were highly significant, with the correlation for ΔGpp in (A) having p < 10-9, and p ≤ 10-5 in (B). For the other correlations, p < 10-15. We note that the correlation values for the sample with complex background are considerably lower, suggesting further scope for improvements in our model.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2823707&req=5

Figure 3: Importance ranking of thermodynamic probe properties including target-side modeling. Importance ranking of thermodynamic probe properties including target-side modelling. The left hand figure (A) shows the Guide ranking for the clearly expressed genes showing no cross-hybridization from Wei et al. [23], whereas the right one shows the results for the Pozhitkov et al. [26] dataset (B). ΔG stands for the effective interaction energy including relevant competing intra- and inter-molecular effects. ΔGp labels the free energy of the probe secondary structure, ΔGt the free energy of the probe binding site secondary structure in the target, ΔGh the free energy of the probe-target duplex, and ΔGpp the free energy of probe-probe dimerization within the same probe feature. 'Tm' stands for the melting temperature. Error bars are from bootstrap re-sampling of 90% of all probes. Below the x-axis, the Spearman rank correlation of predictions to the observed signal intensity is shown. All correlations were highly significant, with the correlation for ΔGpp in (A) having p < 10-9, and p ≤ 10-5 in (B). For the other correlations, p < 10-15. We note that the correlation values for the sample with complex background are considerably lower, suggesting further scope for improvements in our model.

Mentions: The importance of alternative thermodynamic properties in a prediction of probe signal intensity according to the GUIDE algorithm was validated by bootstrap as before. Fig. 3 shows that the effective interaction energy ΔG is by far the best predictor of signal intensity for specific hybridization, with about twice the relative importance score compared to alternative predictors like the melting temperature Tm (left panel). In order to test the generic nature of this result, we also examined a complementary tiling array experiment [26] that probes several ribosomal RNAs (rRNAs) with 25 nt oligonucleotides. Besides providing a test case with much shorter probes and a different target type (RNA instead of DNA), hybridization conditions were simpler in this experiment, featuring uniform target concentrations and no complex background (and thus no cross-hybridization). Table 1 compares the two data sets. Also in this very different experiment, the effective interaction energy ΔG was the best predictor of signal intensity, clearly outperforming alternative predictors like Tm (Fig. 3, right panel).


Hybridization thermodynamics of NimbleGen microarrays.

Mueckstein U, Leparc GG, Posekany A, Hofacker I, Kreil DP - BMC Bioinformatics (2010)

Importance ranking of thermodynamic probe properties including target-side modeling. Importance ranking of thermodynamic probe properties including target-side modelling. The left hand figure (A) shows the Guide ranking for the clearly expressed genes showing no cross-hybridization from Wei et al. [23], whereas the right one shows the results for the Pozhitkov et al. [26] dataset (B). ΔG stands for the effective interaction energy including relevant competing intra- and inter-molecular effects. ΔGp labels the free energy of the probe secondary structure, ΔGt the free energy of the probe binding site secondary structure in the target, ΔGh the free energy of the probe-target duplex, and ΔGpp the free energy of probe-probe dimerization within the same probe feature. 'Tm' stands for the melting temperature. Error bars are from bootstrap re-sampling of 90% of all probes. Below the x-axis, the Spearman rank correlation of predictions to the observed signal intensity is shown. All correlations were highly significant, with the correlation for ΔGpp in (A) having p < 10-9, and p ≤ 10-5 in (B). For the other correlations, p < 10-15. We note that the correlation values for the sample with complex background are considerably lower, suggesting further scope for improvements in our model.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2823707&req=5

Figure 3: Importance ranking of thermodynamic probe properties including target-side modeling. Importance ranking of thermodynamic probe properties including target-side modelling. The left hand figure (A) shows the Guide ranking for the clearly expressed genes showing no cross-hybridization from Wei et al. [23], whereas the right one shows the results for the Pozhitkov et al. [26] dataset (B). ΔG stands for the effective interaction energy including relevant competing intra- and inter-molecular effects. ΔGp labels the free energy of the probe secondary structure, ΔGt the free energy of the probe binding site secondary structure in the target, ΔGh the free energy of the probe-target duplex, and ΔGpp the free energy of probe-probe dimerization within the same probe feature. 'Tm' stands for the melting temperature. Error bars are from bootstrap re-sampling of 90% of all probes. Below the x-axis, the Spearman rank correlation of predictions to the observed signal intensity is shown. All correlations were highly significant, with the correlation for ΔGpp in (A) having p < 10-9, and p ≤ 10-5 in (B). For the other correlations, p < 10-15. We note that the correlation values for the sample with complex background are considerably lower, suggesting further scope for improvements in our model.
Mentions: The importance of alternative thermodynamic properties in a prediction of probe signal intensity according to the GUIDE algorithm was validated by bootstrap as before. Fig. 3 shows that the effective interaction energy ΔG is by far the best predictor of signal intensity for specific hybridization, with about twice the relative importance score compared to alternative predictors like the melting temperature Tm (left panel). In order to test the generic nature of this result, we also examined a complementary tiling array experiment [26] that probes several ribosomal RNAs (rRNAs) with 25 nt oligonucleotides. Besides providing a test case with much shorter probes and a different target type (RNA instead of DNA), hybridization conditions were simpler in this experiment, featuring uniform target concentrations and no complex background (and thus no cross-hybridization). Table 1 compares the two data sets. Also in this very different experiment, the effective interaction energy ΔG was the best predictor of signal intensity, clearly outperforming alternative predictors like Tm (Fig. 3, right panel).

Bottom Line: We demonstrate the benefits of an improved model for microarray hybridization and assess the relative contributions of the probe-target binding strength and the different competing structures.This analysis illustrates the importance of considering both the effects of probe-target binding strength and the different competing structures.Besides their relevance for the design of new arrays, our results demonstrate the value of improving thermodynamic models for the read-out and interpretation of microarray signals.

View Article: PubMed Central - HTML - PubMed

Affiliation: WWTF Chair of Bioinformatics, Boku University Vienna, Muthgasse 18, 1190 Vienna, Austria. Ulrike.Mueckstein@boku.ac.at

ABSTRACT

Background: While microarrays are the predominant method for gene expression profiling, probe signal variation is still an area of active research. Probe signal is sequence dependent and affected by probe-target binding strength and the competing formation of probe-probe dimers and secondary structures in probes and targets.

Results: We demonstrate the benefits of an improved model for microarray hybridization and assess the relative contributions of the probe-target binding strength and the different competing structures. Remarkably, specific and unspecific hybridization were apparently driven by different energetic contributions: For unspecific hybridization, the melting temperature Tm was the best predictor of signal variation. For specific hybridization, however, the effective interaction energy that fully considered competing structures was twice as powerful a predictor of probe signal variation. We show that this was largely due to the effects of secondary structures in the probe and target molecules. The predictive power of the strength of these intramolecular structures was already comparable to that of the melting temperature or the free energy of the probe-target duplex.

Conclusions: This analysis illustrates the importance of considering both the effects of probe-target binding strength and the different competing structures. For specific hybridization, the secondary structures of probe and target molecules turn out to be at least as important as the probe-target binding strength for an understanding of the observed microarray signal intensities. Besides their relevance for the design of new arrays, our results demonstrate the value of improving thermodynamic models for the read-out and interpretation of microarray signals.

Show MeSH