ERC FUNDED PROJECTS

The amazing rate of progress in the computer technologies and telecommunications presents many new challenges for Optimization Theory. New problems are usually very big in size, very special in structure and possibly have a distributed data support. This makes them unsolvable by the standard optimization methods. In these situations, old theoretical models, based on the hidden Black-Box information, cannot work. New theoretical and algorithmic solutions are urgently needed. In this project we will concentrate on development of fast optimization methods for problems of big and very big size. All the new methods will be endowed with provable efficiency guarantees for large classes of optimization problems, arising in practical applications. Our main tool is the acceleration technique developed for the standard Black-Box methods as applied to smooth convex functions. However, we will have to adapt it to deal with different situations. The first line of development will be based on the smoothing technique as applied to a non-smooth functions. We propose to substantially extend this approach to generate approximate solutions in relative scale. The second line of research will be related to applying acceleration techniques to the second-order methods minimizing functions with sparse Hessians. Finally, we aim to develop fast gradient methods for huge-scale problems. The size of these problems is so big that even the usual vector operations are extremely expensive. Thus, we propose to develop new methods with sublinear iteration costs. In our approach, the main source for achieving improvements will be the proper use of problem structure. Our overall aim is to be able to solve in a routine way many important problems, which currently look unsolvable. Moreover, the theoretical development of Convex Optimization will reach the state, when there is no gap between theory and practice: the theoretically most efficient methods will definitely outperform any homebred heuristics.

2 090 038 €

Start date: 2018-09-01, End date: 2023-08-31

Work conducted in my laboratory on the trypanosome killing factor of human serum led to the identification of the primate-specific Apolipoprotein L1 (APOL1) as a novel pore-forming protein with striking similarities with proteins of the apoptotic BCL2 family. APOL1 belongs to a family of proteins induced under inflammatory conditions in myeloid and endothelial cells. APOL1 is efficiently neutralized by the SRA protein of Trypanosoma rhodesiense, accounting for the ability of this trypanosome subspecies to infect humans and cause sleeping sickness. We found that natural APOL1 variants escaping SRA neutralization and therefore conferring human resistance to T. rhodesiense are associated with chronic kidney disease. Moreover, transgenic mice expressing these APOL1 variants exhibit an obese phenotype. Our unpublished results also indicate that APOLs control the lifespan of dendritic cells and podocytes activated by viral stimuli. Therefore, we propose that the pathology of APOL variants is due to their deregulated activity on the control of the cellular lifespan in myeloid/endothelial cells activated by pathogen detection. This project aims at characterizing (i) the molecular mechanism by which APOLs control the lifespan of activated dendritic cells and podocytes, which has direct impact on innate immunity and inflammation, and (ii) the mechanism by which APOL1 variants cause pathology. In addition, we plan to detail the physiological function of APOLs by studying the phenotype of transgenic mice either expressing human APOL1 (wild-type and variants) or devoid of APOL genes, which we have recently generated. Finally, we propose to exploit the extraordinary potential of trypanosomes for antigenic variation in order to produce SRA variants able to neutralize the pathogenic APOL1 variants. Preliminary experiments suggest that in podocytes SRA antagonizes APOL1 induction by viral stimulus and subsequent cell death, opening new perspectives to treat kidney disease.

2 250 000 €

Start date: 2015-09-01, End date: 2021-06-30

The project will address the following key question: How can we provide fibre-like connectivity to moving objects (robots, humans) with the following characteristics: very high dedicated bitrate of 100 Gb/s per object, very low latency of <10 μs, very high reliability of 99.999%, very high density of more than one object per m2 and this at low power consumption? Achieving this would be groundbreaking and it requires a completely new and high-risk approach: applying close proximity wireless communications using low interference ultra-small cells (called “ATTO-cells”) integrated in floors and connected to antennas on the (parallel) floor-facing surface of ground moving objects. This makes it possible to obtain very high densities with very good channel conditions. The technological challenges involved are groundbreaking in mobile networking (overall architecture, handover with extremely low latencies), wireless subsystems (60 GHz substrate integrated waveguide-based distributed antenna systems connected to RF transceivers integrated in floors, low crosstalk between ATTO-cells) and optical interconnect subsystems (simple non-blocking optical coherent remote selection of ATTO-cells, transparent low power 100 Gb/s coherent optical / RF transceiver interconnection using analogue equalization and symbol interleaving to support 4x4 MIMO). By providing this unique communication infrastructure in high density settings, the ATTO concept will not only support the highly demanding future 5G services (UHD streaming, cloud computing and storage, augmented and virtual reality, a range of IoT services, etc.), but also even more demanding services, that are challenging our imagination such as mobile robot swarms or brain computer interfaces with PFlops computing capabilities. This new concept for ultra-high capacity wireless networks will open up many more opportunities in reconfigurable robot factories, intelligent hospitals, flexible offices, dense public spaces, etc.

2 496 250 €

Start date: 2017-01-01, End date: 2021-12-31

Feeding the growing world population under changing climate conditions poses an unprecedented challenge on global agriculture and our current pace to breed new high yielding crop varieties is too low to face the imminent threats on food security. This ERC project proposes a novel crossing scheme that allows for an expeditious evaluation of combinations of potential yield contributing alleles by unifying ‘classical’ breeding with gene-centric molecular biology. The acronym BREEDIT, a word fusion of breeding and editing, reflects the basic concept of combining breeding with multiplex genome editing of yield related genes. By introducing plants with distinct combinations of genome edited mutations in more than 80 known yield related genes into a crossing scheme, the combinatorial effect of these mutations on plant growth and yield will be evaluated. Subsequent rounds of crossings will increase the number of stacked gene-edits per plant, thus increasing the combinatorial complexity. Phenotypic evaluations throughout plant development will be done on our in-house automated image-analysis based phenotyping platform. The nature and frequency of Cas9-mediated mutations in the entire plant collection will be characterised by multiplex amplicon sequencing to follow the efficiency of CRISPR-cas9 genome editing and to identify the underlying combinations of genes that cause beneficial phenotypes (genetic gain). The obtained knowledge on yield regulatory networks can be directly implemented into current molecular breeding programs and the project will provide the basis to develop targeted breeding schemes implementing the optimal combinations of beneficial alleles into elite material. BREEDIT will be a major step forward in integrating basic knowledge on genes with plant breeding and has the potential to provoke a paradigm shift in improving crop yield.

2 474 790 €

Start date: 2019-09-01, End date: 2024-08-31