Utz-Uwe Haus

• Dr. rer. nat.
• Head of EMEA Research Lab at HPE Switzerland

Modeling decision-dependent scenario probabilities in stochastic programs is difficult and typically leads to large and highly non-linear MINLPs that are very difficult to solve. In this paper, we develop a new approach to obtain a compact representation of the recourse function using a set of binary decision diagrams (BDDs) that encode a nested co...

It is well known that (reduced, ordered) binary decision diagrams (BDDs) can sometimes be compact representations of the full solution set of Boolean optimization problems. Recently they have been suggested to be useful as discrete relaxations in integer and constraint programming (Hoda et al. 2010). We show that for every independence system there...

Associating categories with measured or observed attributes is a central challenge for discrete mathematics in life sciences. We propose a new concept to formalize this question: Given a binary matrix of objects and attributes, determine all attribute sets characterizing object sets of cardinality t1t1 that do not characterize any object set of siz...

We call a matrix completely mixable if the entries in its columns can be permuted so that all row sums are equal. If it is not completely mixable, we want to determine the smallest maximal and largest minimal row sum attainable. These values provide a discrete approximation of of minimum variance problems for discrete distributions, a problem motiv...

Robust optimization problems are conventionally solved by reformulation as non-robust problems. We propose a direct method to separate split cuts for robust mixed-integer programs with polyhedral uncertainty sets. The method generalizes the well-known cutting plane procedure of Balas. Computational experiments show that applying cutting planes dire...

A method for efficient scheduling of hybrid classical-quantum workflows is presented, based on standard tools available on common supercomputer systems. Moderate interventions by the user are required, such as splitting a monolithic workflow in to basic building blocks and ensuring the data flow. This bares the potential to significantly reduce idl...

The Workflows Community Summit gathered 111 participants from 18 countries to discuss emerging trends and challenges in scientific workflows, focusing on six key areas: time-sensitive workflows, AI-HPC convergence, multi-facility workflows, heterogeneous HPC environments, user experience, and FAIR computational workflows. The integration of AI and...

OpenCUBE aims to develop an open-source full software stack for Cloud computing blueprint deployed on EPI hardware, adaptable to emerging workloads across the computing continuum. OpenCUBE prioritizes energy awareness and utilizes open APIs, Open Source components, advanced SiPearl Rhea processors, and RISC-V accelerator. The project leverages repr...

Accurate and fast simulation of HPC job scheduling is an important tool for exploring the effect of different scheduling strategies on production systems and for providing insight into future HPC design. Current realistic simulations are computationally intensive and cannot provide a rapid feedback loop to facilitate the development of novel schedu...

In recent years the use of FPGAs to accelerate scientific applications has grown, with numerous applications demonstrating the benefit of FPGAs for high performance workloads. However, whilst High Level Synthesis (HLS) has significantly lowered the barrier to entry in programming FPGAs by enabling programmers to use C++, a major challenge is that m...

Identifying the boundary beyond which quantum machines provide a computational advantage over their classical counterparts is a crucial step in charting their usefulness. Gaussian boson sampling (GBS), in which photons are measured from a highly entangled Gaussian state, is a leading approach in pursuing quantum advantage. State-of-the-art GBS expe...

https://cordis.europa.eu/project/id/764759/results

Among the broad variety of challenges that arise from workloads in a converged HPC and Cloud infrastructure, data movement is of paramount importance, especially oncoming exascale systems featuring multiple tiers of memory and storage. While the focus has, for years, been primarily on optimizing computations, the importance of improving data handli...

Identifying the boundary beyond which quantum machines provide a computational advantage over their classical counterparts is a crucial step in charting their usefulness. Gaussian Boson Sampling (GBS), in which photons are measured from a highly entangled Gaussian state, is a leading approach in pursuing quantum advantage. State-of-the-art quantum...

Traditional compiler optimization theory distinguishes three separate classes of cache miss -- Cold, Conflict and Capacity. Tiling for cache is typically guided by capacity miss counts. Models of cache function have not been effectively used to guide cache tiling optimizations due to model error and expense. Instead, heuristic or empirical approach...

It is well known that (reduced, ordered) binary decision diagrams (BDDs) can sometimes be compact representations of the full solution set of Boolean optimization problems. Recently they have been suggested to be useful as discrete relaxations in integer and constraint programming (Hoda, van Hoeve, and Hooker 2010). We show that for every independe...

Logical models for cellular signaling networks are recently attracting wide interest: Their ability to integrate qualitative information at different biological levels, from receptor-ligand interactions to gene-regulatory networks, is becoming essential for understanding complex signaling behavior. We present an overview of Boolean modeling paradig...

Experimental strategy. (A) Workflow. Mice were trained in a shuttle box to discriminate between rising and falling FM tones. Six or 24 h after avoidance training, mice were sacrificed and striatum (STR), hippocampus (HIP), frontal cortex (FC), and auditory cortex (AC) were removed. Synaptic proteins (PSD-enriched fraction) were collected for each b...

Correlation plots of relative synaptic levels of frontal cortex proteins. Mean abundances relative to NV of frontal cortex proteins monitored 6 h (left) and 24 h (right) after behavioural experiments are plotted on a double logarithmic scale, comparing AV and FS (upper part), AV and TS (middle part), and FS and TS (lower part). Each data point repr...

Analysis of regulated protein networks after aversive learning-summary of all brain areas and time points. Network analysis was performed using Ingenuity Pathway Analysis ™ (IPA). IPA of all proteins with significantly altered levels in the AV group reveals that the majority of these proteins form a complex network. Proteins identified to be regula...

Correlation plots of relative synaptic levels of striatal proteins. Mean abundances relative to NV of striatal proteins monitored 6 h (left) and 24 h (right) after behavioural experiments are plotted on a double logarithmic scale, comparing AV and FS (upper part), AV and TS (middle part), and FS and TS (lower part). Each data point represents a uni...

Correlation plots of relative synaptic levels of auditory cortex proteins. Mean abundances relative to NV of auditory cortex proteins monitored 6 h (left) and 24 h (right) after behavioural experiments are plotted on a double logarithmic scale, comparing AV and FS (upper part), AV and TS (middle part), and FS and TS (lower part). Each data point re...

Analysis of regulated protein networks after aversive learning - comparison of protein regulation in striatum (STR), hippocampus (HIP), frontal cortex (FC), auditory cortex (AC). Proteins found to be down-regulated are indicated in blue. Time points after training (6 h and 24h) are pooled. The protein names are as given in the legend to Figure S7.

MIAPE-compliant MS description

MIAPE-compliant MSI-description Table S3 Mean protein abundances in active groups relative to naïve animal groups.

Cluster analysis was performed using the mathematical software package "DanteR" (Pacific Northwest National Laboratory/http://omics.pnl.gov). Log2 of mean values of relative protein levels (FS/NV, TS/NV and AV/NV) were clustered using the K-means algorithm (K=15) on Euclidean distance metrics without data scaling. Protein accession numbers (Swisspr...

Correlation plots of relative synaptic levels of hippocampal proteins. Mean abundances relative to NV of auditory cortex proteins monitored 6 h (left) and 24 h (right) after behavioural experiments are plotted on a double logarithmic scale, comparing AV and FS (upper part), AV and TS (middle part), and FS and TS (lower part). Each data point repres...

Analysis of regulated protein networks after aversive learning - comparison of protein regulation in the various brain regions at 6 h or 24 h after aversive learning. STR - striatum, HIP - hippocampus, FC - frontal cortex, AC - auditory cortex, down-regulated proteins are indicated in blue. The protein names are as given in the legend to Figure S7.

Changes in synaptic efficacy underlying learning and memory processes are assumed to be associated with alterations of the protein composition of synapses. Here, we performed a quantitative proteomic screen to monitor changes in the synaptic proteome of four brain areas (auditory cortex, frontal cortex, hippocampus striatum) during auditory learnin...

Chemical synapses are highly specialized cell-cell contacts for communication between neurons in the CNS characterized by complex and dynamic protein networks at both synaptic membranes. The cytomatrix at the active zone (CAZ) organizes the apparatus for the regulated release of transmitters from the presynapse. At the postsynaptic side, the postsy...

For a Boolean Matrix A a binary vector v is called t- frequent if Av has at least t entries of value supp (v). Given two parameters t1 < t2 the t1-frequent but t2-infrequent vectors of a matrix represent a Boolean function that has two domains of (opposite) monotonicity. These functions were studied for the purpose of data analysis and abstract con...

Elementary modes (EMs) and minimal cut sets (MCSs) provide important techniques for metabolic network modeling. Whereas EMs describe minimal subnetworks that can function in steady state, MCSs are sets of reactions whose removal will disable certain network functions. Effective algorithms were developed for EM computation while calculation of MCSs...

Motivated by fundamental problems in chemistry and biology we study cluster graphs arising from a set of initial states $S\subseteq\Z^n_+$ and a set of transitions/reactions $M\subseteq\Z^n_+\times\Z^n_+$. The clusters are formed out of states that can be mutually transformed into each other by a sequence of reversible transitions. We provide a sol...

The merged network of TCR/CD4/CD28 and IL-2R signaling. The top layer represents input nodes. The bottom layer represents the output, i.e. molecules including transcription factors that become activated. Solid black arrows indicate activating interactions with a black circle denoting AND-connections. For clarity, activating influences with arrows p...

STAT activation after TCR stimulation of mouse T cells. Primary mouse T cells and mouse T-cell blasts were stimulated as indicated in Protocol S1 and analyzed by Western blotting for the activation of STAT3 and STAT5. Irrelevant lanes have been cut out from the blots. (TIF)

IL-2R signaling does not affect TCR expression. Human T-cell blasts were either stimulated with IL-2 for 30 min or left untreated. The level of TCR (CD3) and IL-2Rα (CD25) surface expression were measured by flow cytometry. Unstained cells are indicated with the broken line, untreated cells are represented with the grey line, and stimulated cells i...

List of network components. The components of the merged TCR-IL-2R network including biological names and interpretation of the ON state. (PDF)

A feasible negative feedback loop. Usually the cycle of FYN, PAG, CSK, and LCKP1 would lead to an infeasible solution as the activating interactions (solid lines) between the four proteins demand that they are in the same state. This is in conflict with the inhibition of LCKP1 by CSK (dotted red line) which requires that one of them is active and t...

Determination of CD25 and Annexin V/PI by FACS staining. After resting for 24 hr and before restimulation with IL-2, T-cell blasts were stained with a FITC-coupled anti-CD25 antibody to determine the percentage of CD25-positive cells (A) or a FITC-coupled Annexin V antibody and propidium iodide to determine the number of viable cells (B). (TIF)

Dose response of IL-2. Human T-cell blasts were stimulated by the indicated dilutions of IL-2 [10,000 U]. Cell lysates were analyzed by Western blotting for phosphorylated ERK and STAT3. β-actin was analyzed as loading control. (TIF)

The reversible SFK inhibitor PP2 does not fully block AKT. The blot is a longer exposure of the blots of Figure 3B and 4B demonstrating that the irreversible PI3K inhibitor WM is more efficient than PP2 in blocking SFK-dependent AKT-phosphorylation following IL-2 stimulation. (TIF)

The TCR/CD4/CD28 signaling network. To a large extent the signaling network was published previously in a different graphical layout [3]. The top layer represents input nodes. The bottom layer represents the output, i.e. molecules including transcription factors that become activated. Solid black arrows indicate activating interactions with a black...

Stimulation of mouse T cells. Method describing the isolation, culture, and stimulation of primary mouse T cells and mouse T-cell blasts. (PDF)

Model merging as an IFFSAT problem. Restating the problem of merging two networks as a projection of a larger network, including a formal proof of the algorithmic complexity. (PDF)

Protocol for Boolean model merging. A step-by-step protocol for the semi-automated merging of Boolean networks. (PDF)

List of clauses for the merged network. The implication formulas for each component of the merged network are provided along with references supporting the interactions. (PDF)

Activation states of the merged network upon different stimuli. Lists the state vectors for the merged network with TCR and IL-2R stimulation alone or in combination. (PDF)

T cells orchestrate the adaptive immune response, making them targets for immunotherapy. Although immunosuppressive therapies prevent disease progression, they also leave patients susceptible to opportunistic infections. To identify novel drug targets, we established a logical model describing T-cell receptor (TCR) signaling. However, to have a mod...

This paper focuses on combinatorial feasibility and optimization problems that arise in the context of parameter identification of discrete dynamical systems. Given a candidate parametric model for a physical system and a set of experimental observations, the objective of parameter identification is to provide estimates of the parameter values for...

We consider a generalization of the unsplittable maximum two-commodity flow problem on undirected graphs where each commodity $i\in{1,2}$ can be split into a bounded number $k_i$ of equally-sized chunks that can be routed on different paths. We show that in contrast to the single-commodity case this problem is NP-hard, and hard to approximate to wi...

A binary matrix has the Consecutive Ones Property (C1P) if its columns can be ordered in such a way that all 1's on each row are consecutive. A Minimal Conflicting Set is a set of rows that does not have the C1P, but every proper subset has the C1P. Such submatrices have been considered in comparative genomics applications, but very little is known...

The determination of all chemical reaction networks composed of elementary reactions for a given net chemical reaction is one of the fundamental problems in chemistry, since the decomposition elucidates the reaction mechanism. It is essential in a wide range of applications: from the derivation of rate laws in physical chemistry to the design of la...

A binary matrix has the Consecutive Ones Property (C1P) if its columns can be ordered in such a way that all 1's on each row are consecutive. A Minimal Conflicting Set is a set of rows that does not have the C1P, but every proper subset has the C1P. Such submatrices have been considered in comparative genomics applications, but very little is known...

A binary matrix has the Consecutive Ones Property (C1P) if its columns can be ordered in such a way that all 1’s on each row are consecutive. A Minimal Conflicting Set is a set of rows that does not have the C1P, but every proper subset has the C1P. Such submatrices have been considered in comparative genomics applications, but very little is known...

Motivated by fundamental problems in chemistry and biology we study cluster graphs arising from a set of initial states $S\subseteq\Z^n_+$ and a set of transitions/reactions $M\subseteq\Z^n_+\times\Z^n_+$. The clusters are formed out of states that can be mutually transformed into each other by a sequence of reversible transitions. We provide a sol...

We propose a static and a dynamic approach to model biological signaling networks, and show how each can be used to answer relevant biological questions. For this, we use the two different mathematical tools of Propositional Logic and Integer Programming. The power of discrete mathematics for handling qualitative as well as quantitative data has so...

The authors have declared that no competing interests exist. This work was supported by the German Federal Ministry of Education and Research (HepatoSys and FORSYS-Centre MaCS (Magdeburg Centre for Systems Biology)), the Ministry of Education and Research of Saxony-Anhalt (Research Center “Dynamic Systems”), and the Helmholtz Alliance on Systems B...

A simple model of signal transduction networks in molecular biology consists of CNF formulas with two and three literals per clause. A necessary condition for correctness of the network model is satisfiability of the formulas. Deciding satisfiability turns out to be NP-complete. However, for a subclass that still is of practical interest, a linear...

In this paper, we present a method to determine globally optimal schedules for cyclically operated plants where activities have to be scheduled on limited resources. In cyclic operation, a large number of entities is processed in an identical time scheme. For strictly cyclic operation, where the time offset between entities is also identical for al...

A simple model of signal transduction networks in molecular biology consists of CNF formulas with two and three literals per clause. A necessary condition for correctness of the network model is satisfiability of the formulas. Deciding satisfiability turns out to be NP-complete. However, for a subclass that still is of practical interest, a linear...

We propose a static and a dynamic approach to model biological signaling networks, and show how each can be used to answer relevant biological questions. For this we use the two different mathematical tools of Propositional Logic and Integer Programming. The power of discrete mathematics for handling qualitative as well as quantitative data has so...

Given a metabolic network in terms of its metabolites and reactions, our goal is to efficiently compute the minimal knock-out sets of reactions required to block a given behavior. We describe an algorithm that improves the computation of these knock-out sets when the elementary modes (minimal functional subsystems) of the network are given. We also...

Polyhedral relaxation is a powerful tool for determining global bounds on optimal solutions in chemical process synthesis. Combined reaction distillation processes are considered as a challenging application example. To reduce complexity of the resulting mixed integer linear optimization problems, model reduction by means of wave functions is propo...

1. Abstract Successful mixed integer nonlinear programming algorithms rely on computing tight over- and under- estimators for nonlinear functions. With the advances of mixed integer linear programming, polyhedral approximations over subregions of the original feasible domain have proven to be a successful tool in recent years. Of course, the smalle...

Given a metabolic network in terms of its metabolites and reactions, our goal is to efficiently compute the minimal knock out sets of reactions required to block a given behaviour. We describe an algorithm which improves the computation of these knock out sets when the elementary modes (minimal functional subsystems) of the network are given. We al...

This paper introduces an exact mathematical approach based on combinatorial optimization to analyze continuous linear countercurrent chromatographic processes. The analysis is based on a given set of input parameters including, in particular, purity requirements and number of plates per zone. As an outcome of the analysis either a proof of infeasib...

Cellular decisions are determined by complex molecular interaction networks. Large-scale signaling networks are currently being reconstructed, but the kinetic parameters and quantitative data that would allow for dynamic modeling are still scarce. Therefore, computational studies based upon the structure of these networks are of great interest. Her...

List of Compounds in the Logical T Cell Model Model name corresponds to the name in Figure 1 and Table S2. Common abbreviations are those usually used in the literature, while name is the whole name. Type classifies the molecules, if applies, as follows: K = Kinase, T = Transcription Factor, P = Phosphatase, A = Adaptor Protein, R = Receptor, G = G...

Hyperarcs of the Logical T Cell Signaling Model (see Figure 1 and Methods) Exclamation mark (“!”) denotes a logical NOT, and dots within the equations indicate AND operations. The names of the substances in the explanations are those used in the model and Figure 1; the biological names are displayed in Table S1. In the case where two pools of a mol...

In this paper a new approach for computing global bounds on optimal solutions of mixed-integer nonlinear programs is presented. These type of problems frequently arise in optimal design of chemical processes. The approach is based on a hierarchy of polyhedral relaxations leading to mixed-integer linear programs, which can be solved rigorously. Appl...

This paper considers the problem of approximating a given crystallite orientation distribution function (codf) by a set of texture components. Problems of this type arise for example if the codf has to be reconstructed from discrete orientations or if one looks for a physical interpretation of the codf. The same problem is encountered if crystallog...

This paper is concerned with computer-aided optimal design of combined reaction−distillation processes. The production of solvent 2,3-dimethylbutene-1 by isomerization of 2,3-dimethylbutene-2 is considered as an innovative benchmark problem. Possible process candidates are a reactive distillation column, a reactor coupled to a nonreactive distillat...

This paper is concerned with the computer-aided optimal design of reaction-distillation processes. The production of solvent 2,3-dimethylbutene-1 by isomerization of 2,3-dimethylbutene-2 is considered as an innovative benchmark problem. Possible process candidates are a reactive distillation column, a reactor coupled to a nonreactive distillation c...

T-cells are able to distinguish foreign antigens among the myriads of self-antigens presented in our body, and organize a directed response against them. While the central sensor for this recognition process is the T-cell receptor (TCR), which binds to peptide-MHC complexes, other signals are also sensed by additional receptors, giving rise to a co...

. This paper introduces an exact primal augmentation algorithm for solving general linear integer programs. The algorithm iteratively substitutes one column in a tableau by other columns that correspond to irreducible solutions of certain linear diophantine inequalities. We prove that various versions of our algorithm are finite. It is a major conc...

We present a new “primal” algorithm for the stable set problem. It is based on a purely combinatorial construction that can transform every graph into a perfect graph by replacing nodes with sets of new nodes. The transformation is done in such a way that every stable set in the perfect graph corresponds to a stable set in the original graph. The a...

Some operations are described that transform every graph into a perfect graph by replacing nodes with sets of new nodes. The transformation is done in such a way that every stable set in the perfect graph corresponds to a stable set in the original graph. These operations can be used in an augmentation procedure for finding a maximum weighted stabl...

this paper we will discuss how to e#ciently apply the Integral Basis Method to arbitrary binary integer programming problems. It turns out that the procedure is most effective if started at a solution quite close to optimality. If combinatorial problems were to be solved, one could apply any problem-specific heuristic to derive such a solution. Our...

An immediate generalization of the classical McKay correspondence for Gorenstein quotient spaces

This paper deals with algorithmic issues related to the design of an augmentation algorithm for general and 0/1-integer programs. We recall the approach of integer pivoting and introduce the family of Gomory-Young augmentation vectors that can be derived from a simplex tableau. Furthermore, a technique of combining Gomory-Young vectors and combinat...

This paper introduces an exact algorithm for solving integer programs, neither using cutting planes nor enumeration techniques. It is a primal augmentation algorithm that relies on iteratively substituting one column by columns that correspond to irreducible solutions of certain linear diophantine inequalities. We demonstrate the algorithm's potent...

An immediate generalization of the classical McKay correspondence for Gorenstein quotient spaces $\Bbb{C}^{r}/G$ in dimensions $r\geq 4$ would primarily demand the existence of projective, crepant, full desingularizations. Since this is not always possible, it is natural to ask about special classes of such quotient spaces which would satisfy the a...