1 / 22

 ratios and fold changes

3000. 3000. x3. ?. 1500. 200. 1000. 0. ?. x1.5. A. A. B. B. C. C. But what if the gene is “off” (below detection limit) in one condition?.  ratios and fold changes. Fold changes are useful to describe continuous changes in expression.  ratios and fold changes.

chenoa
Download Presentation

 ratios and fold changes

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. 3000 3000 x3 ? 1500 200 1000 0 ? x1.5 A A B B C C But what if the gene is “off” (below detection limit) in one condition? ratios and fold changes Fold changes are useful to describe continuous changes in expression

  2. ratios and fold changes The idea of the log-ratio (base 2) 0: no change +1: up by factor of 21 = 2 +2: up by factor of 22 = 4 -1: down by factor of 2-1 = 1/2 -2: down by factor of 2-2 = ¼ A unit for measuring changes in expression: assumes that a change from 1000 to 2000 units has a similar biological meaning to one from 5000 to 10000. What about a change from 0 to 500? - conceptually - noise, measurement precision

  3.  How to compare microarray intensities with each other?  How to address measurement uncertainty (“variance”)?  How to calibrate (“normalize”) for biases between samples? Questions

  4. Systematic Stochastic o similar effect on many measurements o corrections can be estimated from data o too random to be ex-plicitely accounted for o remain as “noise” Calibration Error model Sources of variation amount of RNA in the biopsy efficiencies of -RNA extraction -reverse transcription -labeling -fluorescent detection probe purity and length distribution spotting efficiency, spot size cross-/unspecific hybridization stray signal

  5. bi per-sample normalization factor bk sequence-wise probe efficiency hik multiplicative noise ai per-sample offset eik additive noise  The two component model measured intensity = offset + gain  true abundance

  6. “multiplicative” noise “additive” noise  The two-component model raw scale log scale B. Durbin, D. Rocke, JCB 2001

  7. Parameterization two practically equivalent forms (h<<1)

  8.  variance stabilizing transformations Xu a family of random variables with EXu=u, VarXu=v(u). Define var f(Xu ) independent of u derivation: linear approximation

  9. variance stabilizing transformations f(x) x

  10. 1.) constant variance (‘additive’) 2.) constant CV (‘multiplicative’) 3.) offset 4.) additive and multiplicative  variance stabilizing transformations

  11. the “glog” transformation - - - f(x) = log(x) ———hs(x) = asinh(x/s) P. Munson, 2001 D. Rocke & B. Durbin, ISMB 2002 W. Huber et al., ISMB 2002

  12. generalized log-ratio difference log-ratio variance: constant part proportional part glog raw scale log glog

  13. parameter estimation (vsn package) o maximum likelihood estimator: straightforward – but sensitive to outliers o model is for genes that are unchanged; differentially transcribed genes act as outliers. o robust variant of ML estimator, à la Least Trimmed Sum of Squares regression. o works well as long many genes are not differentially transcribed (<50% throughout the intensity range)

  14. “usual” log-ratio 'glog' (generalized log-ratio) c1, c2are experiment specific parameters (~level of background noise)

  15.  Variance Bias Trade-Off Estimated log-fold-change log glog Signal intensity

  16.  Variance-bias trade-off and shrinkage estimators Shrinkage estimators: a general technology in statistics: pay a small price in bias for a large decrease of variance, so overall the mean-squared-error (MSE) is reduced. Particularly useful if you have few replicates. Generalized log-ratio is a shrinkage estimator for fold change

  17.  “Single color normalization” • n red-green arrays (R1, G1, R2, G2,… Rn, Gn) • within/between slides • for each slide i=1…n • calculate Mi= log(Ri/Gi), Ai= ½ log(Ri*Gi) • normalize Mi vs Ai • Then normalize M1…Mn • all at once • normalize the combined matrix (R, G) • then calculate log-ratios or any other contrast you like

  18.  What about non-linear effects? o Good data operate in the linear regime, where fluorescence intensity increases proportionally to target abundance (see e.g. Affymetrix dilution series) Two reasons for non-linearity: oAt the high intensity end:saturation/quenching. This can and should be avoided experimentally - loss of data! oAt the low intensity end:background offsets,

  19.  Non-linear or affine linear?

  20. Bioinformatics and computational biology solutions using R and Bioconductor, R. Gentleman, V. Carey, W. Huber, R. Irizarry, S. Dudoit, Springer (2005). Variance stabilization applied to microarray data calibration and to the quantification of differential expression. W. Huber, A. von Heydebreck, H. Sültmann, A. Poustka, M. Vingron. Bioinformatics 18 suppl. 1 (2002), S96-S104. Exploration, Normalization, and Summaries of High Density Oligonucleotide Array Probe Level Data. R. Irizarry, B. Hobbs, F. Collins, …, T. Speed. Biostatistics 4 (2003) 249-264. Error models for microarray intensities. W. Huber, A. von Heydebreck, and M. Vingron. Encyclopedia of Genomics, Proteomics and Bioinformatics. John Wiley & sons (2005). Differential Expression with the Bioconductor Project. A. von Heydebreck, W. Huber, and R. Gentleman. Encyclopedia of Genomics, Proteomics and Bioinformatics. John Wiley & sons (2005). References

  21. Anja von Heydebreck (Darmstadt) Robert Gentleman (Seattle) Günther Sawitzki (Heidelberg) Martin Vingron (Berlin) Annemarie Poustka, Holger Sültmann, Andreas Buness, Markus Ruschhaupt (Heidelberg) Rafael Irizarry (Baltimore) Judith Boer (Leiden) Anke Schroth (Heidelberg) Friederike Wilmer (Hilden) Acknowledgements

More Related