Search

SPACE4AI-D tool is an open source design-time to find the optimal solution for component placement and resource selection problem in computing continuum while some response time performance constraints are fulfilled. This tool is a part of AISPRINT project and the source codes witha real system example (video processing) are provided here.

The Deep Learning (DL) paradigm gained remarkable popularity in recent years. DL models are used to tackle increasingly complex problems, making the training process require considerable computational power. The parallel computing capabilities offered by modern GPUs partially fulfill this need, but the high costs related to GPU as a Service solutions in the cloud call for efficient capacity planning and job scheduling algorithms to reduce operational costs via resource sharing. In this work, we jointly address the online capacity planning and job scheduling problems from the perspective of cloud end-users. We present a Mixed Integer Linear Programming (MILP) formulation, and a path relinking-based method aiming at optimizing operational costs by (i) rightsizing Virtual Machine (VM) capacity at each node, (ii) partitioning the set of GPUs among multiple concurrent jobs on the same VM, and (iii) determining a due-date-aware job schedule. An extensive experimental campaign attests the effectiveness of the proposed approach in practical scenarios: costs savings up to 97% are attained compared with first-principle methods based on, e.g., Earliest Deadline First, cost reductions up to 20% are obtained with respect to a previously proposed Hierarchical Method and up to 95% against a dynamic programming-based method from the literature. Scalability analyses show that systems with up to 100 nodes and 450 concurrent jobs can be managed in less than 7 seconds. The validation in a prototype cloud environment shows a deviation below 5% between real and predicted costs.

Loss of CDH1/Cadherin-1 is a common step towards the acquisition of an abnormal epithelial phenotype. In gastric cancer (GC), mutation and/or downregulation of CDH1/Cadherin-1 is recurrent in sporadic and hereditary diffuse GC type. To approach the molecular events downstream of CDH1/Cadherin-1 alterations and their relevance in gastric carcinogenesis, we queried public databases for genetic and DNA methylation data in search of molecular signatures with a still-uncertain role in the pathological mechanism of GC. In all GC subtypes, modulated genes correlating with CDH1/Cadherin-1 aberrations are associated with stem cell and epithelial-to-mesenchymal transition pathways. A higher level of genes upregulated in CDH1-mutated GC cases is associated with reduced overall survival. In the diffuse GC (DGC) subtype, genes downregulated in CDH1-mutated compared to cases with wild type CDH1/Cadherin-1 resulted in being strongly intertwined with the DREAM complex. The inverse correlation between hypermethylated CpGs and CDH1/Cadherin-1 transcription in diverse subtypes implies a common epigenetic program. We identified nonredundant protein-encoding isoforms of 22 genes among those differentially expressed in GC compared to normal stomach. These unique proteins represent potential agents involved in cell transformation and candidate therapeutic targets. Meanwhile, drug-induced and CDH1/Cadherin-1 mutation-related gene expression comparison predicts FIT, GR-127935 hydrochloride, amiodarone hydrochloride in GC and BRD-K55722623, BRD-K13169950, and AY 9944 in DGC as the most effective treatments, providing cues for the design of combined pharmacological treatments. By integrating genetic and epigenetic aspects with their expected functional outcome, we unveiled promising targets for combinatorial pharmacological treatments of GC.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants of concern (VOCs) emerge for their capability to better adapt to the human host aimed and enhance human-to-human transmission. Mutations in spike largely contributed to adaptation. Viral persistence is a prerequisite for intra-host virus evolution, and this likely occurred in immunocompromised patients who allow intra-host long-term viral replication. The underlying mechanism leading to the emergence of variants during viral persistence in the immunocompromised host is still unknown. Here, we show the existence of an ensemble of minor mutants in the early biological samples obtained from an immunocompromised patient and their dynamic interplay with the master mutant during a persistent and productive long-term infection. In particular, after 222 days of active viral replication, the original master mutant, named MB610, was replaced by a minor quasispecies (MB61222) expressing two critical mutations in spike, namely Q493K and N501T. Isolation of the two viruses allowed us to show that MB61222 entry into target cells occurred mainly by the fusion at the plasma membrane (PM), whereas endocytosis characterized the entry mechanism used by MB610. Interestingly, coinfection of two human cell lines of different origin with the SARS-CoV-2 isolates highlighted the early and dramatic predominance of MB61222 over MB610 replication. This finding may be explained by a faster replicative activity of MB61222 as compared to MB610 as well as by the capability of MB61222 to induce peculiar viral RNA-sensing mechanisms leading to an increased production of interferons (IFNs) and, in particular, of IFN-induced transmembrane protein 1 (IFITM1) and IFITM2. Indeed, it has been recently shown that IFITM2 is able to restrict SARS-CoV-2 entry occurring by endocytosis. In this regard, MB61222 may escape the antiviral activity of IFITMs by using the PM fusion pathway for entry into the target cell, whereas MB610 cannot escape this host antiviral response during MB61222 coinfection, since it has endocytosis as the main pathway of entry. Altogether, our data support the evidence of quasispecies fighting for host dominance by taking benefit from the cell machinery to restrict the productive infection of competitors in the viral ensemble. This finding may explain, at least in part, the extraordinary rapid worldwide turnover of VOCs that use the PM fusion pathway to enter into target cells over the original pandemic strain.

Combined antiretroviral therapy (cART) for HIV-1 dramatically slows disease progression among HIV+individuals. Currently, lymphoma represents the main cause of death among HIV-1-infected patients. Detection of p17 variants (vp17s) endowed with B-cell clonogenic activity in HIV-1-seropositive patients with lymphoma suggests their possible role in lymphomagenesis. Here, we demonstrate that the clonogenic activity of vp17s is mediated by their binding to PAR1 and to PAR1-mediated EGFR transactivation through Gq protein. The entire vp17s-triggered clonogenic process is MMPs dependent. Moreover, phosphoproteomic and bioinformatic analysis highlighted the crucial role of EGFR/PI3K/Akt pathway in modulating several molecules promoting cancer progression, including RAC1, ABL1, p53, CDK1, NPM, Rb, PTP-1B, and STAT1. Finally, we show that a peptide (F1) corresponding to the vp17s functional epitope is sufficient to trigger the PAR1/EGFR/PI3K/Akt pathway and bind PAR1. Our findings suggest novel potential therapeutic targets to counteract vp17-driven lymphomagenesis in HIV+patients.

The repository includes the source code and the datasets of SPACE4AI-D which is a design time tool of the AI-SPRINT. The results are included in the AI-SPRINT project deliverable "D2.1 - First release and evaluation of the AI-SPRINT design tools".

This is the source code and dataset that we used in the paper: "A Randomized Greedy Method for AI Applications Component Placement and Resource Selection in Computing Continua" which is published in 2021 IEEE International Conference on Joint Cloud Computing (JCC) Proceedings. What is included in zip file: Two main python files: main.py is the python file related to the first section of experimental results which runs the code related to the paper plots. main_larg_scale.py is the python file related to the second section of the experimental results which run random greedy for some different number of components as reported in the "scalability analysis" subsection. ConfigFiles: this folder includes all necessary input files which are json files. Input_file.json includes input lambda interval and step, the Bandwidth Scenario of network domain 2 (ND2) and the number of iterations in the proposed random greedy algorithm. Random_Greedy.json file include system descriptions related to the random greedy algorithm in which some components are free to place in edge or cloud. OnlyCloud.json file includes system descriptions related to cloud only scenario in which some components can be placed only on cloud VMs. OnlyEdge.json file includes system descriptions related to edge only scenario in which some components can be placed only on edge devices. large_scale: this folder include the system description files for different scales including 5, 10, 15 and 20 components. Output_files: After running the code, the output will be saved in this folder that is include the plots and best solutions for lambdas intervals. The official version of this tool is located in the following link: https://gitlab.polimi.it/ai-sprint/space4ai-d

The pandemic caused by the SARS-CoV-2 has created the need of compounds able to interfere with the biological processes exploited by the virus. Doxycycline, with its pleiotropic effects, including anti-viral activity, has been proposed as a therapeutic candidate for COVID-19 and about twenty clinical trials have started since the beginning of the pandemic.To gain information on the activity of doxycycline against SARS-CoV-2 infection and clarify some of the conflicting clinical data published, we designed in vitro binding tests and infection studies with a pseudotyped virus expressing the spike protein, as well as a clinically isolated SARS-CoV-2 strain. Doxycycline inhibited the transduction of the pseudotyped virus in Vero E6 and HEK-293 T cells stably expressing human receptor angiotensin-converting enzyme 2 but did not affect the entry and replication of SARS-CoV-2.Although this conclusion is apparently disappointing, it is paradigmatic of an experimental approach aimed at developing an integrated multidisciplinary platform. To avoid wasting precious time and resources we believe very stringent experimental criteria are needed in the preclinical phase, including infectious studies with SARS-CoV-2 in the platform before moving on to [failed] clinical trials.Author SummaryThe pandemic caused by the SARS-CoV-2 virus has created a completely unusual situation in rapidly searching for compounds able to interfere with the biological processes exploited by the virus. This new scenario has substantially changed the timing of drug development which has also resulted in the generation of controversial results, proving that the transition from computational screening to the clinical application requires great caution and careful studies. It is therefore necessary to establish new paradigms for evaluating the efficacy of a potential active molecule.We set up a preclinical platform aimed at identifying molecules active against SARS-CoV-2 infection developing a multidisciplinary approach based on very stringent experimental criteria, comprising in-silico studies, in vitro binding tests and infection studies with pseudovirus expressing the spike protein as well as clinically isolated SARS-CoV-2 strains. We focused our attention on doxycycline which has been suggested as potential therapeutic candidate for treating COVID-19 and is currently employed in about twenty clinical trials. Doxycycline resulted effective in inhibiting the transduction of pseudovirus but it did not affect the entry and replication of SARS-CoV-2. The results obtained underline the need to define more stringent and controlled pharmacological approaches before wasting precious time and resources with clinical trials.

Artificial Intelligence (AI) and Deep Learning (DL) algorithms are currently applied to a wide range of products and solutions. DL training jobs are highly resource demanding and they experience great benefits when exploiting AI accelerators (e.g., GPUs). However, the effective management of GPU-powered clusters comes with great challenges. Among these, efficient scheduling and resource allocation solutions are crucial to maximize performance and minimize Data Centers operational costs. In this paper we propose ANDREAS, an advanced scheduling solution that tackles these problems jointly, aiming at optimizing DL training runtime workloads and their energy consumption in accelerated clusters. Experiments based on simulation demostrate that we can achieve a cost reduction between 30 and 62% on average with respect to first-principle methods while the validation on a real cluster shows a worst case deviation below 13% between actual and predicted costs, proving the effectiveness of ANDREAS solution in practical scenarios.

SARS-CoV-2-associated acute respiratory distress syndrome (ARDS) and acute lung injury are life-threatening manifestations of severe viral infection. The pathogenic mechanisms that lead to respiratory complications, such as endothelialitis, intussusceptive angiogenesis, and vascular leakage remain unclear. In this study, by using an immunofluorescence assay and in situ RNA-hybridization, we demonstrate the capability of SARS-CoV-2 to infect human primary lung microvascular endothelial cells (HL-mECs) in the absence of cytopathic effects and release of infectious particles. Preliminary data point to the role of integrins in SARS-CoV-2 entry into HL-mECs in the absence of detectable ACE2 expression. Following infection, HL-mECs were found to release a plethora of pro-inflammatory and pro-angiogenic molecules, as assessed by microarray analyses. This conditioned microenvironment stimulated HL-mECs to acquire an angiogenic phenotype. Proteome analysis confirmed a remodeling of SARS-CoV-2-infected HL-mECs to inflammatory and angiogenic responses and highlighted the expression of antiviral molecules as annexin A6 and MX1. These results support the hypothesis of a direct role of SARS-CoV-2-infected HL-mECs in sustaining vascular dysfunction during the early phases of infection. The construction of virus-host interactomes will be instrumental to identify potential therapeutic targets for COVID-19 aimed to inhibit HL-mEC-sustained inflammation and angiogenesis upon SARS-CoV-2 infection.

HIV-1 matrix protein p17 variants (vp17s) derived from non-Hodgkin's lymphoma (NHL) tissues of HIV-1–seropositive (HIV+) patients promote B-cell growth by activating the Akt signaling pathway. It is fundamental to understand the role played by vp17s in producing a microenvironment that fosters lymphoma development and progression. Therefore, we asked whether vp17s could be secreted from infected cells in their biologically active form. In this study, we show that two B-cell growth-promoting vp17s, NHL-a101 and NHL-a102, characterized by amino acid insertions at position 117 to 118 (Ala–Ala) or 125 to 126 (Gly–Asn), respectively, are secreted from HIV-1–infected Jurkat T cells during the active phase of viral replication. Secretion of biologically active vp17s also occurred in HeLa cells nucleofected with a plasmid expressing the entire Gag gene, following proteolytic cleavage of the Gag precursor polyprotein (Pr55Gag) by cellular aspartyl proteases. Binding of Pr55Gag to phosphatidylinositol-(4,5)-bisphosphate was indispensable for allowing the unconventional secretion of both wildtype p17 and vp17s. Indeed, here we demonstrate that inhibition of Pr55Gag binding to phosphatidylinositol-(4,5)-bisphosphate by using neomycin, or its enzymatic depletion achieved by overexpression of 5ptaseIV, significantly impair the secretion of p17s. We also demonstrated that heparan sulfate proteoglycans were involved in tethering p17s at the cell surface. This finding opens up an interesting way for investigating whether tethered p17s on the surface of HIV-1 reservoirs may represent a likely target for immune-mediated killing.

HIV-1 matrix protein p17 variants (vp17s) derived from non-Hodgkin's lymphoma (NHL) tissues of HIV-1–seropositive (HIV+) patients promote B-cell growth by activating the Akt signaling pathway. It is fundamental to understand the role played by vp17s in producing a microenvironment that fosters lymphoma development and progression. Therefore, we asked whether vp17s could be secreted from infected cells in their biologically active form. In this study, we show that two B-cell growth-promoting vp17s, NHL-a101 and NHL-a102, characterized by amino acid insertions at position 117 to 118 (Ala–Ala) or 125 to 126 (Gly–Asn), respectively, are secreted from HIV-1–infected Jurkat T cells during the active phase of viral replication. Secretion of biologically active vp17s also occurred in HeLa cells nucleofected with a plasmid expressing the entire Gag gene, following proteolytic cleavage of the Gag precursor polyprotein (Pr55Gag) by cellular aspartyl proteases. Binding of Pr55Gag to phosphatidylinositol-(4,5)-bisphosphate was indispensable for allowing the unconventional secretion of both wildtype p17 and vp17s. Indeed, here we demonstrate that inhibition of Pr55Gag binding to phosphatidylinositol-(4,5)-bisphosphate by using neomycin, or its enzymatic depletion achieved by overexpression of 5ptaseIV, significantly impair the secretion of p17s. We also demonstrated that heparan sulfate proteoglycans were involved in tethering p17s at the cell surface. This finding opens up an interesting way for investigating whether tethered p17s on the surface of HIV-1 reservoirs may represent a likely target for immune-mediated killing.

Significance Immune activation and inflammation are predictors of serious non-AIDS events even in virally suppressed HIV-1−infected individuals. This does not apply to HIV-2−infected patients, who experience a form of attenuated HIV-1 disease. Here, we show that the HIV-1 matrix protein 17 (p17) binds to and activates the common beta chain receptor (βCR). The βCR-activating epitope on p17 is expressed on the matrix protein of HIV-1 ancestors but not on that of HIV-2 and its ancestors. Our finding highlights this epitope as a signature tracing the HIV-1 evolutionary trajectory that may have represented a critical step to enhance the HIV-1 ancestors aggressiveness and early human-to-human transmission. Whether this functional epitope actually marks the pathogenic difference between HIV-1 and HIV-2 needs further investigation., The HIV-1 matrix protein p17 (p17) is a pleiotropic molecule impacting on different cell types. Its interaction with many cellular proteins underlines the importance of the viral protein as a major determinant of human specific adaptation. We previously showed the proangiogenic capability of p17. Here, by integrating functional analysis and receptor binding, we identify a functional epitope that displays molecular mimicry with human erythropoietin (EPO) and promotes angiogenesis through common beta chain receptor (βCR) activation. The functional EPO-like epitope was found to be present in the matrix protein of HIV-1 ancestors SIV originated in chimpanzees (SIVcpz) and gorillas (SIVgor) but not in that of HIV-2 and its ancestor SIVsmm from sooty mangabeys. According to biological data, evolution of the EPO-like epitope showed a clear differentiation between HIV-1/SIVcpz-gor and HIV-2/SIVsmm branches, thus highlighting this epitope on p17 as a divergent signature discriminating HIV-1 and HIV-2 ancestors. P17 is known to enhance HIV-1 replication. Similarly to other βCR ligands, p17 is capable of attracting and activating HIV-1 target cells and promoting a proinflammatory microenvironment. Thus, it is tempting to speculate that acquisition of an epitope on the matrix proteins of HIV-1 ancestors capable of triggering βCR may have represented a critical step to enhance viral aggressiveness and early human-to-human SIVcpz/gor dissemination. The hypothesis that the p17/βCR interaction and βCR abnormal stimulation may also play a role in sustaining chronic activation and inflammation, thus marking the difference between HIV-1 and HIV-2 in term of pathogenicity, needs further investigation.

Background Since the first outbreak of SARS-CoV-2, the clinical characteristics of the Coronavirus Disease 2019 (COVID-19) have been progressively changed. Data reporting a viral intra-host and inter-host evolution favouring the appearance of mild SARS-CoV-2 strains are since being accumulating. To better understand the evolution of SARS-CoV-2 pathogenicity and its adaptation to the host, it is therefore crucial to investigate the genetic and phenotypic characteristics of SARS-CoV-2 strains circulating lately in the epidemic. Methods Nasopharyngeal swabs have been analyzed for viral load in the early (March 2020) and late (May 2020) phases of epidemic in Brescia, Italy. Isolation of SARS-CoV-2 from 2 high viral load specimens identified on March 9 (AP66) and on May 8 (GZ69) was performed on Vero E6 cells. Amount of virus released was assessed by quantitative PCR. Genotypic characterization of AP66 and GZ69 was performed by next generation sequencing followed by an in-depth in silico analysis of nucleotide mutations. Results The SARS-CoV-2 GZ69 strain, isolated in May from an asymptomatic healthcare worker, showed an unprecedented capability of replication in Vero E6 cells in the absence of any evident cytopathic effect. Vero E6 subculturing, up to passage 4, showed that SARS-CoV-2 GZ69 infection was as productive as the one sustained by the cytopathic strain AP66. Whole genome sequencing of the persistently replicating SARS-CoV-2 GZ69 has shown that this strain differs from the early AP66 variant in 9 nucleotide positions (C2939T; C3828T; G21784T; T21846C; T24631C; G28881A; G28882A; G28883C; G29810T) which lead to 6 non-synonymous substitutions spanning on ORF1ab (P892S; S1188L), S (K74N; I95T) and N (R203K, G204R) proteins. Conclusions Identification of the peculiar SARS-CoV-2 GZ69 strain in the late Italian epidemic highlights the need to better characterize viral variants circulating among asymptomatic or paucisymptomatic individuals. The current approach could unravel the ways for future studies aimed at analyzing the selection process which favours viral mutations in the human host.

Background: Mixed Cryoglobuminaemia(MC) is a small vessel vasculitis characterized by the presence of a monoclonal rheumatoid factor(RF) and polyclonal immunoglobulins that precipitate at cold temperature, with a large spectrum of clinical manifestations. MC is strongly associated with hepatitis C(HCV) infection>90%. The new direct–acting antivirals(DAA) against HCV, which can eradicate the virus in >95% of patients, are supposed to have benefits on clinical manifestations of MC, nonetheless the amount of data is still limited. Patients with severe vasculitis may require anti-CD20 agent Rituximab(RTX) over DAA. However, despite the strong rationale, the advantage of dual therapy is still debated and the optimal therapeutic schedule unknown. Objectives: To investigate the impact of DAA treatment on clinical manifestations and biochemical parameters of disease activity in HCV-related MC patients who reached complete viral eradication, and if pre-treatment with RTX was able to define a subset of patients with a different outcome. Methods: Data was collected from 31 patients with MC, mean age 68 yrs, female 80%, mean disease duration 10 yrs who were treated with DAA and reached complete viral eradication, at 3 time points: before starting DAA (pre-DDA), after completing DAA (post-DAA), 52 weeks after completing DAA (52w). Additionally, a population of 9 patients with the same disease (mean age 65 yrs, female 77%, mean disease duration 9 yrs) but treated with one or more cycles of RTX prior receiving DAA, was analysed at the same timepoints and additionally before starting RTX(pre-RTX). At each time point clinical data (presence/absence of purpura, neuropathic manifestations, arthralgia, fatigue, nephritis) and biochemical parameters (% cryoglobulins, C3, C4 and RF) were recorded. The change in clinical and biochemical parameters at different timepoints were evaluated. Results: Among patients treated with DAA only, the prevalence of clinical manifestations before starting treatment was the following: purpura 31%,neuropathic manifestations 52%,arthralgia 28%,fatigue 14%, nephritis 3%. We observed a significant reduction in purpura after treatment (pre-DDA vs post-DAA=31% vs 13%,p=0.044) that persisted at 52w (pre-DAA vs 52w=31% vs 0%,p=0.004%). This in parallel with a significant reduction in the level of cryoglobulins (pre-DAA vs 52w=3.99±5.28 vs 1.6 ±1.7,p=0.029 and pre-DAA vs 52w=3.99 ± 5.28 vs 0.92 ± 1.75,p=0.008). Among patients pre-treated with RTX, the prevalence of clinical manifestations before starting RTX was: purpura 78%, neuropathic manifestations 100%,arthralgia 44%,fatigue 50%,nephritis 22%;and before starting DAA was: purpura 22%,neuropathic manifestations 100%,arthralgia 0%,fatigue 0%,nephritis 0%. A significant reduction in the rate of purpura (pre-RTX vs pre-DAA=78% vs 22%, p=0.009%) and arthralgia (44% vs 0%,p=0.012%) after RTX treatment was observed, but no difference in any clinical parameter when comparing pre-RTX with post-DAA and 52w. No significant change in any of the serological parameter was observed. Conclusion: DAA treatment is associated with rapid and sustained decrease in the level of cryoglobulins and related clinical improvement, specifically purpura. Patients who had been pre-treated with RTX showed clinical improvement in terms of purpura and arthralgia, albeit transient; subsequent treatment with DAA did not result in any immediate or long term additional clinical benefit. References [1] Mazzaro C. Clin Exp Rheum2017. Disclosure of Interests: Maria Di Cicco: None declared, Davide Antonio Filippini: None declared, Federica Filippini: None declared, Laura Belloli: None declared, Valeria Campanella: None declared, Cinzia Casu: None declared, Michel Chevallard: None declared, Marina Muscara: None declared, Mariaeva Romano Speakers bureau: janssen, Emanuela Schito: None declared, Elisa Verduci: None declared, Massimo Puoti: None declared, Maria Vinci: None declared, Oscar Massimiliano Epis Speakers bureau: BMS

The aim of the study was to evaluate the sensitization to metals in patients with Co-Cr hip prosthesis. Peripheral blood mononuclear cells (PBMC) were collected from 14 healthy donors and three groups of patients: 10 candidates for primary total joint replacements, 11 patients with well-fixed implant and 13 patients with aseptic loosening of the hip prosthesis. PBMCs were cultured with the metal ions employed for implant manufacturing and the expression of CD69 activation antigen on CD3/T lymphocytes was detected by flow cytometry. Chromium extract increased significantly the expression of CD3/CD69 phenotype in patients with loosening of hip prosthesis. The chromium-induced 'activation index' was higher in patients with loosening of hip prosthesis than in healthy donors and in pre-implant patients. The cobalt-stimulated PBMC of patients with either well-fixed or loosened prosthesis had an 'activation index' significantly higher than healthy donors. The activation index values were used to graduate the PBMC-response as 'normal' (or = 0.9 and2), 'low' (0.9) and 'high' (or = 2): an high-activation index was observed only in chromium-exposed PBMC of patients with prosthesis. Our data show that chromium released from orthopedic implants could be responsible for the lymphocyte sensitization and flow cytometry is an easy and reliable method for monitoring the hypersensitivity state in patients with metal prostheses. Activated lymphocytes in the peri-implant tissue are likely to elicit a localized immune response and contribute to maintain the inflammatory process evolving in the implant failure.

The authors evaluated the ability of bone cement to modify the pro"le of pro-in#ammatory cytokines secreted by the immune cells. Peripheral blood mononuclear cells (PBMC) collected from healthy individuals were cultured with cement extracts and tested to assess the release of IL-1b, TNFa, GM-CSF and IL-6 in both unstimulated and PHA-stimulated PBMC. The cytokine release of unstimulated PBMC was very poor, and in particular the IL-1b was undetectable: the addition of cement extract increased both TNFa and GM-CSF release and decreased IL-6, sometimes signi"cantly. The most recurrent observation in PHA-stimulated PBMCs exposed to bone cement extract was the increase in both IL-1b and IL-6 release, while both the mean concentration and the index of release of TNFa and GM-CSF were changeable. In conclusion our results showed that leachable components of some bone cements can induce in vitro the release of pro-in#ammatory cytokines which are known to be involved in the bone resorption associated with aseptic loosening of hip prostheses. These "ndings allowed us to identify materials endowed with the highest in#ammatory power. ( 2000 Elsevier Science Ltd. All rights reserved.

The biocompatibility of two cyanoacrylate surgical glues (Glubran and Glubran 2), supplied by General Enterprise Marketing, Viareggio, Lucca, Italy, was tested through cytotoxicity and blood compatibility tests and the evaluation of antimicrobial activity. Cytotoxicity and blood compatibility tests were performed on the polymerized glues. Using the neutral red uptake test, the extracts from Glubran and Glubran 2 after polymerization were non-toxic to L929 cells only when diluted 1: 10 with culture medium. Glubran and Glubran 2 induced a significant decrease of activated partial thromboplastin time (APTT), which is favourable with regard to the desired haemostasis. The APTT shortening determines a haemostatic effect and therefore contribute to the tissue adhesion induced by the glues. Otherwise, no significant variation of prothrombin activity, fibrinogen, platelet number, total and differential leukocyte count was induced by the glues, which, in addition, did not show haemolytic effect. There was no difference between Glubran and Glubran 2 regarding haemocompatibility. The antimicrobial ability of the unpolymerized glues was tested onto Bacillus subtilis var. niger for 3 weeks: neither Glubran nor Glubran 2 were found effective in this respect. In conclusion, we can assume that cytotoxicity was severe with the undiluted glues, but was acceptable when glues were diluted. On the contrary, blood compatibility was acceptable for the intended use of the glues. No difference was found between Glubran and Glubran 2 after polymerization.

The ability of bone cements to modify the apoptotic program in activated immune cells and the mechanisms by which they act were evaluated. Mononuclear cells were collected from healthy individuals, cultured for 4 and 24 h with phytohemoagglutinina-P and cement extracts and then tested to assess: (a) cell viability; (b) early apoptotic events, by Annexin V/propidium iodide staining; and (c) the expression of pro- (p53, c-myc, ICE) and anti-apoptotic (bcl-2) genes. After 4 h three cements were able to increase significantly the percentage of apoptotic cells, while after 24 h no differences were found. The proportion of dead cells was not significantly changed at either culture time. The simultaneous expression of both pro-apoptotic (ICE, c-myc, p53) and antiapoptotic genes (bcl-2) was investigated only with regard to the materials which induced significant changes in apoptosis: two cements induced the p53 expression, while the third down-regulated bcl-2. As apoptosis regulates the balance of immune response, the authors recommend that the interaction between materials and immune cells should be assessed, so that the use of pro-apoptotic materials may be avoided in patients with immune defects.

The aim of this study was the evaluation of the release of bone-resorbing cytokines in peripheral blood mononuclear cells (PBMC) of patients with aseptic loosening of Co–Cr hip prostheses. TNF-α, IL-6, and GM-CSF were measured in both unstimulated and PHA-stimulated PBMC, and in PBMC cultured in the presence of chromium and cobalt extracts. Blood samples from healthy donors were used as controls. Serum samples of both patients and healthy donors were tested to determine the concentration of metal ions. The proportion of lymphocyte, monocyte and lympocyte subpopulation in cultured PBMC did not differ in patients and the control group. In unstimulated PBMC the release of TNF was significantly higher in patients than in the control group, while IL-6 was significantly decreased and no change was observed for GM-CSF. When the PBMC were challenged with chromium extract, all the ‘index of cytokine release’ resulted higher in patients than in the control group; cobalt extract increased both the TNF and GM-CSF index, but not the index of IL-6 release. Metal concentrations in serum from patients were significantly increased and correlated with the TNF release in PBMC stimulated with both metal extract. Our results suggest that a CoCr-implant releases a large amount of metal ions which could mediate the priming or the renewal of a cell-mediated hypersensitivity reaction. The prevalence of circulating lymphocytes responsible for the delayed hypersensitivity, namely Th1, would justify both the significant increase of TNF and the significant decrease of IL6 in unstimulated PBMC of patients, as well as the significant increase of the ‘index of cytokine release’ after the challenge with metal ions.

Methyl-tert-butyl ether (MTBE) was submitted to long-term carcinogenicity bioassays on Sprague-Dawley rats. The test compound was delivered in olive oil by stomach tube (gavage), at the doses of 1000, 250, and 0 mg/kg b.w. to groups of 60 males and 60 females, once daily, 4 times weekly, for 104 weeks. All animals were kept under control until spontaneous death. MTBE was found to cause in males an increased incidence of Leydig cell testicular tumors in the group treated with the higher dose, and in females a dose-related increase of leukemias, an increase of dysplastic proliferations of lymphoreticular tissues, and also an increase of uterine sarcomas at the lower tested dose. On the basis of the presented data, MTBE must be considered a potential carcinogen.

Deep learning (DL) methods have recently gained popularity. Training this class of models is, however, computing-intensive, and frequently GPUs are used to boost performance. Although the costs of GPU-based systems are gradually reducing due to the high demand, they are still prohibitive: in public clouds, GPU-powered virtual machines (VMs) time unit price is 5-8x higher than CPU-only VMs. While the cloud remains the most cost-effective and flexible deployment, operation costs can be reduced, in large settings, by rightsizing and sharing resources among multiple processes. This work addresses the online joint capacity planning and job scheduling with due dates problem and proposes alternative matheuristic solution methods. Our objective is to optimize operation costs by: i) rightsizing the VM capacities at each node, ii) partitioning the set of GPUs among multiple concurrent jobs on the same VM, and iii) determining a due-date-aware job schedule. The effectiveness of the proposed hierarchical approach, coupled with an appropriate Mixed Integer Linear Programming formulation, is validated against first-principle methods by relying on simulation. The experiments prove that the efficiency of GPU-based systems evaluated in terms of costs can be improved by 50-70%. Finally, scalability analyses show that the proposed approach enables to solve problem instances with up to 100 nodes in less than one minute on average, making it suitable for practical scenarios.

This deliverable, which is the first release of the market analysis and forecast report for AI-SPRINT, provides a detailed assessment of the market status-quo and market forecast for edge computing with a specific focus on AI and other enabling technologies. It also gives a deep dive into the market forecast of the three main use-cases of the AI-SPRINT project. The European enterprise edge market has grown significantly in the last couple of years, with new solutions developed to address specific industries and use cases as more organisations have started to understand the benefits and opportunities behind edge and other emerging technologies. The COVID-19 pandemic has reshuffled the investment plans of many organisations in the world, in some cases stimulating the need for edge to ensure business continuity, to improve the decision-making process and to support remote operations. European enterprise edge spending is expected to reach €39.6 billion by 2025, driven by the tremendous role of edge in enabling low latency use cases, deployments in locations that are either remote or without IT support staff, and supporting new vendor propositions and a supply chain that is being built up. Major findings in this document include the following: European enterprise edge spending will grow from €18 billion in 2020 to €39.6 billion in 2025, with a five-year Compound Annual Growth Rate (CAGR) of 18%, outpacing the total ICT market average that is forecast to grow at single digits over the same period. Compared with Central and Eastern Europe (CEE), Western Europe sees a more established and developed edge scenario. Nevertheless, edge adoption in CEE is expected to increase over the forecast period, especially within key verticals. Within the three main technology groups, services (including professional and provisioned services) will continue to make half of the total market spending. Hardware will outpace the market average, reaching a fourth of the market, as new hybrid edge infrastructures drive innovations at the edge. Edge supports a wide array of other emerging technology markets referred to as domains, that are driving the overall edge market and generating value outside core data centers. These are Internet of Things (IoT) as the largest domain within the European edge market, followed by artificial intelligence (AI) with tremendous capabilities on real-time data processing, augmented reality (AR)/virtual reality (VR), robotics and drones. The distribution and services sector along with manufacturing and resources have the largest share of European enterprise edge spending. Despite the pandemic heavily hitting some industries within those sectors, edge use cases linked to the Internet of Things (IoT) and AI domains will drive developments, especially in edge environments where edge bridges the gap between IT and operations technologies (OT). The report double clicks into the market dimension for edge solutions by type of edge technology, and by industry, looking also at the AI domain in the context of AI-SPRINT. In particular, the report provides a detailed analysis on the three use cases in the AI- SPRINT context: maintenance and inspection, personalised healthcare and farming 4.0, looking also at the drivers and the barriers for the edge market overall and in the context of specific industry use cases. In the market context chapter, the report presents the significant market developments for the edge market but also an analysis of how the ecosystem is a pivotal point for deploying edge solutions, particularly in the context of the three AI-SPRINT cases., {"references":["Project deliverable","Open Access"]}

The aim of the AI-SPRINT project is to implement a design and runtime framework to accelerate the development of AI applications whose components are spread across the edge-cloud computing continuum. AI-SPRINT tools will allow trading-off application performance (in terms of end-to-end latency or throughput), energy efficiency, and AI models accuracy while providing security and privacy guarantees. This document, in particular, describes the first release and evaluation of the design tools of the AI-SPRINT platform, while the second release and evaluation are due at M24. The first release includes programming abstractions that hide the communications across components and transparently implement the parallelization of the compute-intensive part of the application; quality annotations to enrich applications with constraints on performance and security; performance models to support AI components execution time prediction (both for inference and training tasks); AI Models Network Architecture Search providing solutions to enable developers with limited Machine Learning (ML) expertise to train high-quality models specific to their needs also in terms of Quality of Service (QoS) requirements; Applications design space exploration tools to evaluate multiple alternative candidate deployments for complex applications involving many components. The main results of this release include: The implementation of a prototype application for the AI-SPRINT Healthcare use case that adopts a parallel version of the Cascade Support Vector Machine (CSVM) algorithm to train a model that classifies a dataset to identify atrial fibrillations possibly related to the heart stroke. This implementation adopts the programming model of AI-SPRINT using PyCOMPSs and the dislib distributed ML library, reaching an average f1-score of 0.667 in discriminating atrial fibrillation from normal recordings using the time-frequency features of a balanced selection of the PhysioNet CinC Challenge 2017 available data with shuffled 80/20 training and test sets splits repeated 10 times. The design of QoS annotations corresponding to performance and security constraints. The AISPRINT API will include parsing tools which will extract the decorated functions from the code of the application and produce semi-automatically1 the input required by the AI-SPRINT toolchain (including design, deployment and monitoring). Tools for the definition of performance models and profiling tools based on a ML library, that allows to predict the execution time of inference and training jobs, through feature selection, and hyperparameter tuning. A component for automatic neural architecture search (POPNAS) to automate the search for the best Deep Neural Network (DNN) architecture for a given classification/regression task. Also leveraging the performance models, the algorithm searches for the best network configuration to achieve the higher accuracy in the lowest possible time by searching the Pareto front of the time-accuracy trade-off. A tool (SPACE4AI-D) for the automatic design space exploration in order to minimize the execution cost of the AI application while providing response time guarantees. The output of this tool determines the optimal component placement, resource selection and the optimal number of nodes/VMs which helps the developer to find the optimal placement. For each of the above-mentioned components, an evaluation section is also provided to discuss preliminary results on the adopted technologies and to evaluate the integration of the components according to the project’s milestones., {"references":["Project deliverable","Open Access"]}

The aim of the AI-SPRINT “Artificial intelligence in Secure PRIvacy-preserving computing coNTinuum” project is to develop a platform composed of design and runtime management tools to seamlessly design, partition and operate Artificial Intelligence (AI) applications among the current plethora of cloud-based solutions and AI-based sensor devices (i.e., devices with intelligence and data processing capabilities), providing resource efficiency, performance, data privacy, and security guarantees. The aim of this deliverable is to review the state-of-the-art in techniques used in the development and operation of AI applications in computing continua and the related technologies. The deliverable starts with an overview of AI applications and edge computing market trends. It examines the standardisation and opensource landscape, where many of the emerging standards are mainly focused on the network layer, also highlighting opportunities for open-source community involvement to expand technology availability. Then, the deliverable provides a background on AI applications design, also considering some advanced design trends (e.g., Network Architecture Search, Federated Learning, Deep Neural Networks partitioning) introducing the software development frameworks and hardware solutions which allow to run such applications in a computing continuum. We then report an overview of the state-of-the-art of the solutions offered by the main cloud market players and the European providers. We also extensively discuss existing solutions for applications deployment, monitoring, runtime management, and scheduling considering the emerging Function as a Service paradigm. In the last part of the deliverable, we report an overview of the performance modelling solutions, security, and privacy problems for AI applications in edge environments. It is found that some areas, such as component placement and design space exploration with privacy preservation and performance guarantees, are fairly under-developed, since solutions tailored for AI applications are lacking, and thus offer an opportunity for innovation., {"references":["Project deliverable","Open Access"]}

This document, developed by the AI-SPRINT project, represents an initial version of the platform Architecture deliverable (the final due at M24) that offers an overview of the architectural design. The focus of this document is the identification of different components and interfaces within the AI-SPRINT architectural building blocks, as well as the different strategies necessary for a successful deployment. The role of each of the blocks is defined, as well as the required interfaces for the communication between blocks. The architectural choices are built taking into account the different types of requirements imposed by the applications and services running in the framework through the analysis of the three use cases specified in the project proposal and detailed in AI-SPRINT Deliverable D1.2 - Requirements Analysis., {"references":["Project deliverable","Open Access"]}

L’articolo illustra gli effetti del muro di separazione/segregazione nella vita quotidiana e nelle dinamiche conflittuali che intercorrono tra Israele e Palestina. Il muro come simbolo di incomunicabilità, distanza, de-umanizzazione. Ma quel muro non è “l’ultima parola”... Alcune testimonianze riporteranno lo spiraglio della speranza, perché il dialogo è sempre possibile anche in situazioni dolorose e complesse.

Introduzione al tema degli immigrati di seconda generazione.