Florian Schmid

Genetic Factors of the Disease Course After Sepsis: Rare Deleterious Variants Are Predictive

Sepsis is a life-threatening organ dysfunction caused by dysregulated host response to infection. For its clinical course, host genetic factors are important and rare genomic variants are suspected to contribute. We sequenced the exomes of 59 Greek and 15 German patients with bacterial sepsis divided into two groups with extremely different disease courses. Variant analysis was focusing on rare deleterious single nucleotide variants (SNVs). We identified significant differences in the number of rare deleterious SNVs per patient between the ethnic groups. Classification experiments based on the data of the Greek patients allowed discrimination between the disease courses with estimated sensitivity and specificity > 75%. By application of the trained model to the German patients we observed comparable discriminatory properties despite lower population-specific rare SNV load. Furthermore, rare SNVs in genes of cell signaling and innate immunity related pathways were identified as classifiers discriminating between the sepsis courses. Sepsis patients with favorable disease course after sepsis, even in the case of unfavorable preconditions, seem to be affected more often by rare deleterious SNVs in cell signaling and innate immunity related pathways, suggesting a protective role of impairments in these processes against a poor disease course.

Interpretable Classifiers in Precision Medicine: Feature Selection and Multi-class Categorization

Growing insight into the molecular nature of diseases leads to the definition of finer grained diagnostic classes. Allowing for better adapted drugs and treatments this change also alters the diagnostic task from binary to multi-categorial decisions. Keeping the corresponding multi-class architectures accurate and interpretable is currently one of the key tasks in molecular diagnostics.

Sputnik: ad hoc distributed computation

MOTIVATION In bioinformatic applications, computationally demanding algorithms are often parallelized to speed up computation. Nevertheless, setting up computational environments for distributed computation is often tedious. Aim of this project were the lightweight ad hoc set up and fault-tolerant computation requiring only a Java runtime, no administrator rights, while utilizing all CPU cores most effectively. RESULTS The Sputnik framework provides ad hoc distributed computation on the Java Virtual Machine which uses all supplied CPU cores fully. It provides a graphical user interface for deployment setup and a web user interface displaying the current status of current computation jobs. Neither a permanent setup nor administrator privileges are required. We demonstrate the utility of our approach on feature selection of microarray data. AVAILABILITY AND IMPLEMENTATION The Sputnik framework is available on Github http://github.com/sysbio-bioinf/sputnik under the Eclipse Public License. CONTACT hkestler@fli-leibniz.de or hans.kestler@uni-ulm.de SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

GiANT: gene set uncertainty in enrichment analysis

UNLABELLED Over the past years growing knowledge about biological processes and pathways revealed complex interaction networks involving many genes. In order to understand these networks, analysis of differential expression has continuously moved from single genes towards the study of gene sets. Various approaches for the assessment of gene sets have been developed in the context of gene set analysis (GSA). These approaches are bridging the gap between raw measurements and semantically meaningful terms.We present a novel approach for assessing uncertainty in the definition of gene sets. This is an essential step when new gene sets are constructed from domain knowledge or given gene sets are suspected to be affected by uncertainty. Quantification of uncertainty is implemented in the R-package GiANT. We also included widely used GSA methods, embedded in a generic framework that can readily be extended by custom methods. The package provides an easy to use front end and allows for fast parallelization. AVAILABILITY AND IMPLEMENTATION The package GiANT is available on CRAN. CONTACTS hans.kestler@leibniz-fli.de or hans.kestler@uni-ulm.de.

Semantic ulti- lassifier ystems for the nalysis of ene xpression rofiles

The analysis of biomolecular data from high-throughput screens is typically characterized by the high dimensionality of the measured profiles. Development of diagnostic tools for this kind of data, such as gene expression profiles, is often coupled to an interest of users in obtaining interpretable and low-dimensional classification models; as this facilitates the generation of biological hypotheses on possible causes of a categorization. Purely data driven classification models are limited in this regard. These models only allow for interpreting the data in terms of marker combinations, often gene expression levels, and rarely bridge the gap to higher-level explanations such as molecular signaling pathways. Here, we incorporate into the classification process, additionally to the expression profile data, different data sources that functionally organize these individual gene expression measurements into groups. The members of such Ludwig Lausser · Matthias Platzer · Hans A. Kestler Leibniz Institute on Aging, Jena, Germany [ludwig.lausser,matthias.platzer,hans.kestler]@leibniz-fli.de Mikko J. Sillanpää University of Oulu, Finland mikko.sillanpaa@oulu.fi Ludwig Lausser, Florian Schmid, Hans A. Kestler Medical Systems Biology, Ulm University, Germany [ludwig.lausser,florian-1.schmid,hans.kestler]@uni-ulm.de ⇤ contributed equally ⇤⇤ corresponding author ARCHIVES OF DATA SCIENCE, SERIES A DOI 10.5445/KSP/1000058747/09 KIT SCIENTIFIC PUBLISHING ISSN 2363-9881 Vol. 1, No. 1, S. 157–176, 2016 A E P M C S G 158 L. Lausser, F. Schmid, M. Platzer, MJ. Sillanpää, HA. Kestler a group of measurements share a common property or characterize a more abstract biological concept. These feature subgroups are then used for the generation of individual classifiers. From the set of these classifiers, subsets are combined to a multi-classifier system. Analysing which individual classifiers, and thus which biological concepts such as pathways or ontology terms, are important for classification, make it possible to generate hypotheses about the distinguishing characteristics of the classes on a functional level.

Rank-based classifiers for extremely high-dimensional gene expression data

Predicting phenotypes on the basis of gene expression profiles is a classification task that is becoming increasingly important in the field of precision medicine. Although these expression signals are real-valued, it is questionable if they can be analyzed on an interval scale. As with many biological signals their influence on e.g. protein levels is usually non-linear and thus can be misinterpreted. In this article we study gene expression profiles with up to 54,000 dimensions. We analyze these measurements on an ordinal scale by replacing the real-valued profiles by their ranks. This type of rank transformation can be used for the construction of invariant classifiers that are not affected by noise induced by data transformations which can occur in the measurement setup. Our 10

Three Transductive Set Covering Machines

\times

Ordinal Prototype-Based Classifiers

× 10 fold cross-validation experiments on 86 different data sets and 19 different classification models indicate that classifiers largely benefit from this transformation. Especially random forests and support vector machines achieve improved classification results on a significant majority of datasets.