IPCOS (Belgium)

Abstract Patients with homozygous SMARCA4 loss of function (LOF) aberrations represent a high unmet medical need. Occurring in ∼5% of non-small-cell-lung-cancer (NSCLC) patients, SMARCA4 deficiency represents a distinct molecular subset that is mutually exclusive with most other oncogenic drivers and associated with resistance to radiotherapy, chemotherapy and PD-L1 treatment. SMARCA4 LOF aberrations can co-occur with KRAS-G12C oncogenic mutations in NSCLC, leading to resistance to KRAS targeted therapies. Interestingly, loss of SMARCA4 activity results in a complete dependency on SMARCA2, creating a unique synthetically lethal relationship that has been validated by both genetic and pharmacological approaches. Though developing selective SMARCA2 inhibitors has been challenging due its similarity with SMARCA4, selectivity over SMARCA4 is essential. Multiple Phase 1 clinical trials exploring the SMARCA4/2 dual inhibitor FHD-286 in different indications were discontinued, flagging safety concerns. Thus, there is a clear need to develop a highly potent & selective SMARCA2 small molecule inhibitor to fully unlock the therapeutic potential of this target. At Onco3R Therapeutics, our patient centric approach, integrating deep translational science with rational, structure-based and AI-augmented drug design, has led to the identification of novel SMARCA2 selective small molecule inhibitor series with a best-in-class potency and selectivity profile. Our lead series inhibits SMARCA2 activity with sub-nanomolar potencies while exhibiting over 35-fold selectivity against SMARCA4 in a KRT80 qRT-PCR assay using SMARCA2 or SMARCA4 isogenic cells. Anti-proliferative activity was confirmed in SMARCA4 deficient SK-MEL5 cells. Moreover, Onco3R series exhibit favourable in vitro ADME-safety properties and good PK profiles, enabling robust in vivo characterization. Oral dosing achieved sufficient target coverage to lead to tumor regression in a SMARCA4 deficient RERF-LC-A1 based cell line derived xenograft (CDX) mouse model. Here, we identified unique SMARCA2 selective inhibitor series with a best-in-class potency and selectivity profile. Thorough characterization and optimization efforts are ongoing to identify clinical development candidates with the ultimate goal of achieving greater efficacy along with enhanced safety, providing meaningful benefits to patients diagnosed with SMARCA4 deficient cancer. Citation Format: Lijs Beke, Sandrine Grosse, Shaun Martin, Godelieve Lammens, Pieter Peeters, Bart Stoops, Sandrine Vendeville, Stéphane de Cesco, Pierre Raboisson, Sara Musch, David Moreno Delgado, Line Oste, Francois Gonzalvez. Discovery of novel SMARCA2 small molecule inhibitors with best-in-class potency and selectivity for the treatment of SMARCA4-mutant cancers [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2025 Oct 22-26; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2025;24(10 Suppl):Abstract nr C011.

Abstract Upstream production optimization is concerned with optimizing the entire hydrocarbon value chain from reservoir to sales. In this quest, managing the process facilities to maximize productivity is a critical part. Traditionally, operating companies have addressed it by running offline scenarios to develop a playbook or build physics-based dynamic simulation models. However, a typical offshore facility undergoes significant changes to its operations during its life cycle, which makes the model management tedious and expensive. Few attempts have been made to build data-driven digital twins (machine learning) but they often lack the ability to provide explainable models and lack physical insights. In this work, we propose to use a fast, hybrid, self-learning dynamic process modeling method from routine plant measurements that can be used for real-time forecasts, scenario modeling, process optimization and control. A reduced order modeling method based on input-output dynamic mode decomposition (ioDMD) has been adapted to develop a dynamic process model based on historical data collected from plant sensors. First, we benchmarked the proposed approach with a dynamic simulation model (commercial simulator) using a designed input sequence for training. The ioDMD model simplifies the physical mechanisms to a low dimensional form. Next, we applied the method to an actual offshore deepwater facility based on plant measurements. In both cases, the ioDMD method provided very good predictions without any human intervention. Unlike black-box data-driven methods, the ioDMD method uses an interpretable approach that can be used to explain causal relationships. Observability and controllability of the proposed model can also be easily understood. The proposed ioDMD method provides a unique and sustainable way to combine advanced analytics and physics to develop an explainable dynamic model for the process facility that can be effectively used to assist operations in optimizing performance. The lightweight model lends itself naturally to fast computation that are required for optimization and process control (including IoT edge devices).

This article assesses the potential contribution of economic experiments to evidence‐based policy making in the field of agriculture, with a special focus on the European Union (EU)'s Common Agricultural Policy (CAP). CAP evaluation mostly relies on standard tools such as farm and market simulation models, calibrated with EU‐wide statistical data; statistical and econometric analysis of survey data; and a range of qualitative methods such as interviews with stakeholders, focus group, or Internet‐based public consultation. Yet, the CAP has changed considerably over the past decades, requiring adaptations of its evaluation toolbox. A detailed review of existing studies using economic experiments for designing and evaluating agricultural policies provides the backbone for a comprehensive assessment of the complementarity of experimental approaches with standard evaluation tools. Based on conclusions drawn from a workshop organized with experts, academics, and policy makers of Directorate‐General for Agriculture and Rural Development of the European Commission, the article further provides recommendations aiming at facilitating inclusion of economic experiments into the CAP evaluation toolbox.Cet article étudie la contribution potentielle des méthodes expérimentales pour l'élaboration de politiques agricoles fondée sur des données probantes, en considérant en particulier le cas de la politique agricole commune de l'UE (PAC). L'évaluation de la PAC repose principalement sur des outils standard tels que des modèles de simulation d'exploitations et de marchés agricoles, calibrés avec des données récoltées à l'échelle de l'UE; des analyses statistiques et économétriques sur des données d'enquête; et une série de méthodes qualitatives telles que des entretiens avec les parties prenantes, des groupes de discussion ou des consultations publiques sur Internet. Pourtant, la PAC a considérablement changé ces dernières décennies, ce qui nécessite d'adapter ses méthodes d'évaluation. Un examen détaillé des études existantes mobilisant les approches expérimentales pour la conception et l'évaluation de politiques agricoles permet d'analyser la complémentarité des méthodes expérimentales avec les outils d'évaluation standards. A partir des conclusions d'un atelier organisé avec des experts, des universitaires et des décideurs politiques de la Direction générale de l'agriculture et du développement rural de la Commission européenne, cet article propose en outre des recommandations pour faciliter l'intégration des approches expérimentales dans les méthodes d'évaluation de la PAC.

Abstract The general goal of the present study is to analyze whether children with siblings lag behind their only‐child counterparts in terms of health and nutrition, cognition and educational performance, and non‐cognitive outcomes. We draw on a dataset containing 25 871 observations constructed from three school‐level surveys spanning four provinces in China. The analysis compares children with siblings and only children aged 9 to 14 years old in terms of eight different health, cognitive and non‐cognitive indicators. We find that with the exception of the anemia rate, health outcomes of children with siblings are statistically indistinguishable from those of only children. In terms of cognition, children with siblings performed better than only children. Moreover, outcomes of children with siblings are statistically indistinguishable from those of only children in terms of the non‐cognitive outcomes provided by measures of anxiety. According to our results, the same general findings are true regardless of whether the difference between children with and without siblings is disaggregated by gender.

Abstract This paper describes the successful application of Real-Time Optimization by Shell Malaysia E&P on the Integrated Gas Production System in Sarawak, implementing models for real-time monitoring and optimization of wells and facilities on a gas production network spanning more than 100 wells on more than 40 platforms across a number of different Production Sharing Contracts. We highlight how Digital Oil Field practices enable field-based data to be turned into information, support decision making, and lead to actions that ensure production is optimized continuously. The technology described in this paper is applied to achieve consistent gas supply to meet demand, maximize revenue, and enable improved and timely operational decisions - striking a balance between short- and long-term value, and taking into account the reality of commercial and contractual constraints, finance, and economics. The optimization is data-driven and covers more than 1,000 variables and features multiple, mutually dependent objectives and constraints. The solution has proven significantly better than prior physical model-based solutions, which deliver optimized field settings, but with inherently unstable results, and not fast enough for application in a real-time decision making environment. Field trials have proven a result of: increased condensate production at current or improved expected Ultimate Recovery, whilst maintaining a stable gas supply, fulfilling quality constraints and contractual LNG nominations. This is one of the first successful attempts to implement truly-real-time optimization in a production environment of this size and complexity, including a complicated set of commercial and contractual constraints, and striking a transparent balance between short-term and long-term value. Having proven that a multi-departmental reality can be successfully captured and modeled, might mark the start of a transformation towards embedding intelligent energy to its true potential.

This paper discusses the concepts of linguistic integrity and interpretability. The concepts are used as a framework to design an algorithm to construct linguistic fuzzy models from Numerical Data. The constructed model combines prior knowledge (if present) and numerical information. Two algorithms are presented in this chapter. The main algorithm is the Autonomous Fuzzy Rule Extractor with Linguistic Integrity (AFRELI). This algorithm is complemented with the use of the Fu Zion algorithm created to merge consecutive membership functions while guar anteeing the distinguishability between fuzzy sets. Examples of function approximations and modeling of industrial data are presented as application examples.KeywordsMembership FunctionFuzzy SystemRule BaseRecursive Little SquareLinguistic LabelThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.