The Annals of Applied Statistics

Automated analysis of quantitative image data using isomorphic functional mixed models, with application to proteomics data

Jeffrey S. Morris, Veerabhadran Baladandayuthapani, Richard C. Herrick, Pietro Sanna, and Howard Gutstein

Full-text: Open access


Image data are increasingly encountered and are of growing importance in many areas of science. Much of these data are quantitative image data, which are characterized by intensities that represent some measurement of interest in the scanned images. The data typically consist of multiple images on the same domain and the goal of the research is to combine the quantitative information across images to make inference about populations or interventions. In this paper we present a unified analysis framework for the analysis of quantitative image data using a Bayesian functional mixed model approach. This framework is flexible enough to handle complex, irregular images with many local features, and can model the simultaneous effects of multiple factors on the image intensities and account for the correlation between images induced by the design. We introduce a general isomorphic modeling approach to fitting the functional mixed model, of which the wavelet-based functional mixed model is one special case. With suitable modeling choices, this approach leads to efficient calculations and can result in flexible modeling and adaptive smoothing of the salient features in the data. The proposed method has the following advantages: it can be run automatically, it produces inferential plots indicating which regions of the image are associated with each factor, it simultaneously considers the practical and statistical significance of findings, and it controls the false discovery rate. Although the method we present is general and can be applied to quantitative image data from any application, in this paper we focus on image-based proteomic data. We apply our method to an animal study investigating the effects of cocaine addiction on the brain proteome. Our image-based functional mixed model approach finds results that are missed with conventional spot-based analysis approaches. In particular, we find that the significant regions of the image identified by the proposed method frequently correspond to subregions of visible spots that may represent post-translational modifications or co-migrating proteins that cannot be visually resolved from adjacent, more abundant proteins on the gel image. Thus, it is possible that this image-based approach may actually improve the realized resolution of the gel, revealing differentially expressed proteins that would not have even been detected as spots by modern spot-based analyses.

Article information

Ann. Appl. Stat., Volume 5, Number 2A (2011), 894-923.

First available in Project Euclid: 13 July 2011

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Bayesian analysis false discovery rate functional data analysis functional mixed models functional MRI image analysis isomorphic transformations proteomics 2D gel electrophoresis wavelets


Morris, Jeffrey S.; Baladandayuthapani, Veerabhadran; Herrick, Richard C.; Sanna, Pietro; Gutstein, Howard. Automated analysis of quantitative image data using isomorphic functional mixed models, with application to proteomics data. Ann. Appl. Stat. 5 (2011), no. 2A, 894--923. doi:10.1214/10-AOAS407.

Export citation


  • Ahmed, S. H. and Koob, G. F. (1998). Transition from moderate to excessive drug intake: Change in hedonic set point. Science 282 298–300.
  • Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. Roy. Statist. Soc. Ser. B 57 289–300.
  • Candes, E. J. and Donoho, D. L. (2000). Curvelets, multiresolution representation, and scaling laws. In SPIE Wavelet Applications in Signal and Image Processing VIII (A. Aldroubi, A. F. Laine and M. A. Unser, eds.). Proc. SPIE 4119. Hindawi, New York.
  • Clark, B. N. and Gutstein, H. B. (2008). The myth of automated, high-throughput two-dimensional gel electrophoresis. Proteomics 8 1197–1203.
  • Clyde, M., Parmigiani, G. and Vidakovic, B. (1998). Multiple shrinkage and subset selection in wavelets. Biometrika 85 391–401.
  • Dawid, A. P. (1981). Some matrix-variate distribution theory: Notational considerations and a Bayesian application. Biometrika 68 265–274.
  • Diggle, P. J. and Al Wasel, I. (1997). Spectral analysis of replicated biomedical time series. J. Roy. Statist. Soc. Ser. C 46 31–71.
  • Do, M. N. and Vetterli, M. (2001). Beyond Wavelets (J. Stoeckler and G. V. Welland, eds.). Academic Press, New York.
  • Do, M. N. and Vetterli, M. (2005). The contourlet transform: An efficient directional multiresolution image representation. IEEE Trans. Image Process. 14 2091–2106.
  • Dowsey, A. W., Dunn, M. J. and Yang, G. Z. (2008). Automated image alignment for 2D gel electrophoresis in a high-throughput proteomics pipeline. Bioinformatics 24 950–957.
  • Faergestad, E. M., Rye, M., Walczak, B., Gidskehaug, L., Wold, J. P., Grove, H., Jia, X., Hollung, K., Indahl, U. G., Westad, F., van den Berg, F. and Martens, H. (2007). Pixel-based analysis of multiple images for the indentification of changes: A novel approach applied to unravel patterns of 2-D electrophoresis gel images. Proteomics 7 3450–3461.
  • Feilner, M., Van De Ville, D. and Unser, M. (2005). An orthogonal family of quincunx wavelets with continuously adjustable order. IEEE Trans. Signal Process. 14 499–510.
  • Gygi, S. P., Corthals, G. L., Zhang, Y., Rochon, Y. and Aebersold, R. (2000). Evaluation of two-dimensional gel electrophoresis-based proteome analysis technology. Proc. Natl. Acad. Sci. USA 97 9390–9395.
  • Guo, W. (2002). Functional mixed effects models. Biometrics 58 121–128.
  • Heinrichs, S. C., Menzaghi, F., Schulteis, G., Koob, G. F. and Stinus, L. (1995). Suppression of corticotropin-releasing factor in the amygdala attenuates aversive consequences of morphine withdrawal. Behavioral Pharmacology 6 74–80.
  • Herrick, R. C. and Morris, J. S. (2006). Wavelet-based functional mixed model analysis: Computational considerations. In Proceedings, Joint Statistical Meetings, ASA Section on Statistical Computing 2051–2053. Amer. Statist. Assoc., Alexandria, VA.
  • Karp, N. A. and Lilley, K. S. (2005). Maximizing sensitivity for detecting changes in protein expression: Experimental design using minimal CyDyes. Proteomics 5 3105–3115.
  • Kokkinidis, L., Zacharko, R. M. and Predy, P. A. (1980). Post-amphetamine depression of self-stimulation responding from the substantia nigra: Reversal by tricyclic antidepressants. Pharmacol Biochem. Behav. 12 379–383.
  • Leith, N. J. and Barrett, R. J. (1976). Amphetamine and the reward system: Evidence for tolerance and post-drug depression. Psychopharmacologia 46 19–25.
  • Lilley, K. S. (2003). Protein profiling using two-dimensional difference gel electrophoresis (2-D DIGE). In Current Protocols in Protein Science, Chapter 22, Unit 22.2. Wiley, New York.
  • Mallat, S. G. (1989). A theory for multiresolution signal decomposition: The wavelet representation. IEEE Trans. Pattern Anal. Mach. Intell. 11 674–693.
  • Markou, A. and Koob, G. (1992). Bromocriptine reverses the elevation in intracranial self-stimulation thresholds observed in a rat model of cocaine withdrawal. Neuropsychopharmacology 7 213–224.
  • Morris, J. S. (2010). Supplement to “Automated analysis of quantitative image data using isomorphic functional mixed models, with application to proteomics data.” DOI: 10.1214/10-AOAS407SUPPA, DOI: 10.1214/10-AOAS407SUPPB, DOI: 10.1214/10-AOAS407SUPPC, DOI: 10.1214/10-AOAS407SUPPD.
  • Morris, J. S., Arroyo, C., Coull, B., Ryan, L. M., Herrick, R. and Gortmaker, S. L. (2006). Using wavelet-based functional mixed models to characterize population heterogeneity in accelerometer profiles: A case study. J. Amer. Statist. Assoc. 101 1352–1364.
  • Morris, J. S., Brown, P. J., Herrick, R. C., Baggerly, K. A. and Coombes, K. R. (2008). Bayesian analysis of mass spectrometry data using wavelet-based functional mixed models. Biometrics 12 479–489.
  • Morris, J. S. and Carroll, R. J. (2006). Wavelet-based functional mixed models. J. Roy. Statist. Soc. Ser. B 68 179–199.
  • Morris, J. S., Clark, B. N. and Gutstein, H. B. (2008). Pinnacle: A fast, automatic and accurate method for detecting and quantifying protein spots in 2-dimensional gel electrophoresis data. Bioinformatics 24 529–536.
  • Morris, J. S., Vannucci, M., Brown, P. J. and Carroll, R. J. (2003). Wavelet-based nonparametric modeling of hierarchical functions in colon carcinogenesis. J. Amer. Statist. Assoc. 98 573–583.
  • Morris, J. S., Clark, B. N., Wei, W. and Gutstein, H. B. (2010). Evaluting the performance of new approaches to spot quantification and differential expression in 2-dimensional gel electrophoresis studies. Journal of Proteome Research 9 595–604.
  • Mueller, P., Parmigiani, G., Robert, C. and Rousseau, J. (2004). Optimal sample size for multiple testing: The case of gene expression microarrays. J. Amer. Statist. Assoc. 99 990–1001.
  • Parsons, L. H., Koob, G. F. and Weiss, F. (1995). Serotonin dysfunction in the nucleus accumbens of rats during withdrawal after unlimited access to intravenous cocaine. Journal of Pharmacology and Experimental Therapeutics 274 1182–1191.
  • Ramsay, J. O. and Silverman, B. W. (1997). Functional Data Analysis. Springer, New York.
  • Reiss, P. T. and Ogden, R. T. (2009). Functional generalized linear models with images as predictors. Biometrics. Published online May 8, 2009.
  • Richter, R. and Weiss, F. (1999). In vivo CRF release in rat amygdala is increased during cocaine withdrawal in self-administering rats. Synapse 32 254–261.
  • Sampson, P. D. and Guttorp, P. (1992). Nonparametric estimation of nonstationary spatial covariance structure. J. Amer. Statist. Assoc. 87 108–119.
  • Schulteis, G., Markou, A., Gold, L. H., Stinus, L. and Koob, G. F. (1994). Relative sensitivity to naloxone of multiple indices of opiate withdrawal: A quantitative dose–response analysis. Journal of Pharmacology and Experimental Therapeutics 271 1391–1398.
  • Smith, M. and Fahrmeir, L. (2007). Spatial Bayesian variable selection with application to functional magnetic resonance imaging. J. Amer. Statist. Assoc. 102 417–431.
  • Storey, J. D. (2003). The positive false discovery rate: A Bayesian interpretation and the q-value. Ann. Statist. 31 2013–2035.
  • Strimmer, K. (2008). fdrtool: A versitile R package for estimating local and tail area-based false discovery rates. Bioinformatics 24 1461–1462.
  • Weiss, F., Markou, A., Lorang, M. T. and Koob, G. F. (1992). Basal extracellular dopamine levels in the nucleus accumbens are decreased during cocaine withdrawal after unlimited-access self-ad ministration. Brain Research 593 314–318.

Supplemental materials

  • Supplementary material A: Computational details for wavelet-space implementation of ISO-FMM for image data. Computational details for wavelet implementation of the ISO-FMM for image data, including empirical Bayes method for estimating regularization parameters, MCMC details and Metropolis–Hastings details for covariance parameters.
  • Supplementary material B: Supplementary figures. Supplementary figures, including a virtual 2d gel simulated from the model, a demonstration of the spatial covariance structure induced by the model and 8 plots containing zoomed-in results from analysis of application data in certain interesting regions of the gel.
  • Supplementary material C: Spatial covariance structure in image WFMM. Basic illustration of spatial covariance structure induced by ISO-FMM with 2D wavelet transforms and independence assumed in the wavelet space. Basic demonstration described, and some plots provided. Movie file spatial_covariance.wvm also available as supplementary material to further illustrate these results.
  • Supplementary material D: Movie file illustrating spatial covariance structure of ISO-WFMM with 2D wavelet transform. Windows movie file illustrating the nonstationary spatial covariance structure induced by the ISO-FMM with 2D wavelet bases, with independence assumed among wavelet coefficients. Description of data yielding this movie is provided in the file “Spatial Covariance Structure in Image WFMM.pdf,” also available as supplementary material.