Tammo Rukat
Statistics - Machine Learning - Genomics | DPhil Student @ University of Oxford

# journal papers

• Bayesian Boolean Matrix Factorisation
T. Rukat, C.C. Holmes, M. Titsias, C. Yau. Proceedings of the 34th International Conference on Machine Learning. 2017.

Boolean matrix factorisation aims to decompose a binary data matrix into an approximate Boolean product of two low rank, binary matrices: one containing meaningful patterns, the other quantifying how the observations can be expressed as a combination of these patterns. We introduce the OrMachine, a probabilistic generative model for Boolean matrix factorisation and derive a Metropolised Gibbs sampler that facilitates efficient parallel posterior inference. On real world and simulated data, our method outperforms all currently existing approaches for Boolean matrix factorisation and completion. This is the first method to provide full posterior inference for Boolean Matrix factorisation which is relevant in applications, e.g. for controlling false positive rates in collaborative filtering and, crucially, improves the interpretability of the inferred patterns. The proposed algorithm scales to large datasets as we demonstrate by analysing single cell gene expression data in 1.3 million mouse brain cells across 11 thousand genes on commodity hardware.
  @InProceedings{rukat2017_bayes-boolean,
title = 	 {Bayesian Boolean Matrix Factorisation},
author = 	 {Tammo Rukat and Chris C. Holmes and Michalis K. Titsias and Christopher Yau},
booktitle = 	 {Proceedings of the 34th International Conference on Machine Learning},
pages = 	 {2969--2978},
year = 	 {2017},
editor = 	 {Doina Precup and Yee Whye Teh},
volume = 	 {70},
series = 	 {Proceedings of Machine Learning Research},
address = 	 {International Convention Centre, Sydney, Australia},
month = 	 {06--11 Aug},
publisher = 	 {PMLR},
pdf = 	 {http://proceedings.mlr.press/v70/rukat17a/rukat17a.pdf},
url = 	 {http://proceedings.mlr.press/v70/rukat17a.html},
abstract = 	 {Boolean matrix factorisation aims to decompose a binary data matrix into an approximate Boolean product of two low rank, binary matrices: one containing meaningful patterns, the other quantifying how the observations can be expressed as a combination of these patterns. We introduce the OrMachine, a probabilistic generative model for Boolean matrix factorisation and derive a Metropolised Gibbs sampler that facilitates efficient parallel posterior inference. On real world and simulated data, our method outperforms all currently existing approaches for Boolean matrix factorisation and completion. This is the first method to provide full posterior inference for Boolean Matrix factorisation which is relevant in applications, e.g. for controlling false positive rates in collaborative filtering and, crucially, improves the interpretability of the inferred patterns. The proposed algorithm scales to large datasets as we demonstrate by analysing single cell gene expression data in 1.3 million mouse brain cells across 11 thousand genes on commodity hardware.}
}
• Dynamic contrast-enhanced MRI in mice: an investigation of model parameter uncertainties.
T. Rukat, S. Walker-Samuel, and S. A. Reinsberg. Magnetic Resonance in Medicine, 2014.

PURPOSE:
To establish the experimental factors that dominate the uncertainty of hemodynamic parameters in commonly used pharmacokinetic models.
METHODS:
By fitting simulation results from a multiregion tissue exchange model (Multiple path, Multiple tracer, Indicator Dilution, 4 region), the precision and accuracy of hemodynamic parameters in dynamic contrast-enhanced MRI with four tracer kinetic models is investigated. The impact of various injection rates as well as imprecise knowledge of the arterial input functions is examined.
RESULTS:
Fast injections are beneficial for K(trans) precision within the extended Tofts model and within the two-compartment exchange model but do not affect the other models under investigation. Biases from errors in the arterial input functions are mostly consistent in size and direction for the simple and the extended Tofts model, while they are hardly predictable for the other models. Errors in the hematocrit introduce the greatest loss in parameter accuracy, amounting to an average K(trans) bias of 40% for a 30% overestimation throughout all models.
CONCLUSION:
This simulation study allows the detailed inspection of the isolated impact from various experimental conditions on parameter uncertainty. Because parameter uncertainty comparable to human studies was found, this study represents a validation of preclinical dynamic contrast-enhanced MRI for modeling human tumor physiology.
@article{rukat2015_dynam-mri,
author =	 {Rukat, T. and Walker-Samuel, S and Reinsberg, S. A.},
doi =		 {10.1002/mrm.25319},
issn =	 {1522-2594},
journal =	 {Magnetic Resonance in Medicine},
pages =	 {1979-1987},
title =	 {Dynamic Contrast-Enhanced {MRI} in Mice: An Investigation of Model Parameter Uncertainties},
volume =	 73,
year =	 2015}
• Chain-length dependent growth dynamics of n-alkanes on silica investigated by
energy-dispersive x-ray reflectivity in situ and in real-time.

C. Weber, C. Frank, S. Bommel, T. Rukat, W. Leitenberger, P. Schäfer, F. Schreiber, and S. Kowarik.
The Journal of Chemical Physics. 2012.

We compare the growth dynamics of the three n-alkanes C(36)H(74), C(40)H(82), and C(44)H(90) on SiO(2) using real-time and in situ energy-dispersive x-ray reflectivity. All molecules investigated align in an upright-standing orientation on the substrate and exhibit a transition from layer-by-layer growth to island growth after about 4 monolayers under the conditions employed. Simultaneous fits of the reflected intensity at five distinct points in reciprocal space show that films formed by longer n-alkanes roughen faster during growth. This behavior can be explained by a chain-length dependent height of the Ehrlich-Schwoebel barrier. Further x-ray diffraction measurements after growth indicate that films consisting of longer n-alkanes also incorporate more lying-down molecules in the top region. While the results reveal behavior typical for chain-like molecules, the findings can also be useful for the optimization of organic field effect transistors where smooth interlayers of n-alkanes without coexistence of two or more molecular orientations are required.
@article{weber2012_chain,
author =	 {Weber, C. and Frank, C. and Bommel, S. and Rukat,
T. and Leitenberger, W. and Sch{\"a}fer, P. and
Schreiber, F. and Kowarik, S.},
doi =		 {10.1063/1.4719530},
issn =	 {1089-7690},
journal =	 {The Journal of chemical physics},
number =	 20,
pages =	 204709,
title =	 {Chain-Length Dependent Growth Dynamics of N-Alkanes
on Silica Investigated By Energy-Dispersive X-Ray
Reflectivity in Situ and in Real-Time.},
url =		 {http://www.ncbi.nlm.nih.gov/pubmed/22667583},
volume =	 136,
year = 2012,}