No abstract

No abstract

This paper focuses on the parallelization of an ocean model applying current multicore processor-based cluster architectures to an irregular computational mesh. The aim is to maximize the efficiency of the computational resources used. To make the best use of the resources offered by these architectures, this parallelization has been addressed at all the hardware levels of modern supercomputers: firstly, exploiting the internal parallelism of the CPU through vectorization; secondly, taking advantage of the multiple cores of each node using OpenMP; and finally, using the cluster nodes to distribute the computational mesh, using MPI for communication within the nodes. The speedup obtained with each parallelization technique as well as the combined overall speedup have been measured for the western Mediterranean Sea for different cluster configurations, achieving a speedup factor of 73.3 using 256 processors. The results also show the efficiency achieved in the different cluster nodes and the advantages obtained by combining OpenMP and MPI versus using only OpenMP or MPI. Finally, the scalability of the model has been analysed by examining computation and communication times as well as the communication and synchronization overhead due to parallelization.

This article proposes a new algorithm for parallel synchronous simulation of VHDL designs to be executed on desktop computers. Besides executing VHDL processes in parallel, the algorithm focuses on parallelizing the simulation kernel with special emphasis on signal grouping while maintaining language semantics. Synchronous approaches are the most suitable for shared memory multiprocessor (SMP) desktop computers but may be difficult to parallelize because of the low activity detected in most of the designs. The degree of parallelism is increased in this approach by performing an exhaustive VHDL signal dependencies analysis and avoiding any sequential phase in the simulator. VHDL semantics impose a synchronization barrier after each phase, that is, the process and the kernel simulation phase, as the language definition does not allow simultaneous execution of kernel and processes. These barriers have been relaxed in order to increase the level of parallelism and obtain better performance. Another aspect the new algorithm takes into account is to improve load balancing and locality of references, both critical issues in synchronous simulators, by introducing a new load balancing algorithm that exploits the cyclic characteristics of circuit simulators. These developments make the algorithm suitable for commodity hardware, that is, SMP that are currently used as desktop personal computers.

The physical interaction between metastasis-initiating cells and the pre-existing capillary network (a process known as vascular co-option) is critical during the initial stages of multi-organ metastasis in cancer. As such, this process might provide an opportunity to prevent metastasis. As part of the process of vascular co-option, we observed that brain metastatic cells in the perivascular niche temporarily enter into a novel cell state characterized by a decreased proliferation before resuming their aggressive growth to colonize the organ. Transcriptomic analysis of co-opting metastatic cells confirmed downregulation of MYC signatures, mitotic cell cycle and increased stemness properties. By focusing on one of the top upregulated transcription factors in co-opting cells: MXD4, a MYC antagonist; we have been able to dissect the relevance of this cellular state, that we termed proliferative pause, both respect to the maintenance of the interaction with the vasculature and to the ability to generate macrometastases. As such, targeting MXD4 in lung adenocarcinoma and triple-negative breast cancer metastatic models reduced multi-organ metastases to a level that translates into increased overall survival. Mechanistically, this obliged proliferative pause is linked to a cellular response to the increasing environmental pressure involved in organ colonization. For instance, we found that crossing the blood-brain barrier induced an increased DNA damage due to mechanical constrains leading to nuclear deformation. However, the MXD4-dependent proliferative pause during vascular co-option allows metastatic cells to repair this damage to continue the colonization process. Given the enormous potential to prevent metastasis and our findings dissecting the proliferative pause status, we developed a therapeutic strategy to target vascular co-opting cells. As part of the molecular profile of co-opting cells, we validated their high dependency on Bcl proteins. As such, we have used a Bcl-2 inhibitor (obatoclax) permeable to the blood-brain barrier to target these cells in preventive scenarios. Beyond the preventive therapeutic assays in vivo, we have applied additional clinically-relevant models where preventive strategies could easily translate into the clinical practice. As such, obatoclax post-surgery provided a survival benefit by preventing relapse, as the cells left behind after the local therapy are vascular co-opting cells. Furthermore, our national network of brain metastasis (RENACER) provided us with fresh neurosurgeries and, in a limited cohort of 10 surgeries with extended resections, we were able to identify invasive fronts with metastatic cells co-opting the vasculature. The use of obatoclax to target these cancer cells, which are the seeds of relapse post-surgery, confirmed that targeting vascular co-option could be a novel strategy to prevent metastasis in a clinically relevant situation. Citation Format: Pedro García-Gómez, Diana Retana, Pablo Sanz-Martínez, Irene Salgado-Crespo, Carolina Hernández-Oliver, Maria Isabel García, Oliva Sánchez, Kevin Troulé-Lozano, Verona Villar-Cerviño, Miguel Lafarga-Coscojuela, Fátima Al-Shahrour, RENACER Red Nacional de Metástasis Cerebral, Manuel Valiente. Metastatic colonization requires a proliferative pause linked to vascular co-option [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3516.