Development of an automatised platform for drug discovery in relevant human tissue models

对研究背景的简要描述，包括潜在的影响众所周知，药物发现是一个漫长而昂贵的过程，成功的结果非常低。整个过程通常分三个阶段进行：i)药物分子的合成，ii)临床前动物试验和iii)人体临床试验。临床前动物试验在过去几年中显示出严重的局限性，这是由于动物和人类之间巨大的物种间差距造成的。这往往会导致动物试验结果的广泛不可靠性，导致药物研发项目的巨大损失。通过保留患者特有的特征并更可靠地再现生理和病理场景，人类来源的体外/体外模型的使用打开了该领域的新视野。生物和临床数据的可获得性的增强导致了更复杂的计算模型和生物信息学应用的发展。机器学习模型在治疗应用中显示出巨大的潜力，因为它们可以通过处理大量不同种类的数据集来提供对药物疗效的新的预测性见解。正如相关监管机构(如FDA、EMA)所表示的那样，新的体外-体内耦合硅胶模型的组合已被描述为改善当前药物开发管道的关键方向。在肝病领域，建立新的治疗策略的紧迫性显而易见，包括具有不同病因的广泛病理因素(例如，病毒感染、酗酒、广泛的高脂肪饮食等)。原位肝移植仍然被认为是治疗终末期肝病的金标准。然而，这一做法仍然受到关键负面因素的影响，特别是长期免疫抑制治疗的必要性和目前适合移植的器官短缺。在几种病理情况下(例如，非酒精性脂肪性肝炎-NASH和胆道闭锁)，治疗方案的戏剧性缺乏进一步加剧了这种情况，这对识别和开发新的有效治疗方法提出了全球挑战。目的和目标博士项目致力于开发一种工程过程，旨在从患者的手术样本中推导和测试多种可靠的体外模型(例如，器官样体、精确切割的组织切片)。这些模型将通过使用计算模型和信息学资源，进一步用于评估与标准护理治疗相比的新药候选的疗效。研究方法的新颖性该项目旨在建立一种多学科的方法，通过采用多方面的观点来解决目前药物发现过程中的缺陷。该活动包括三个主要感兴趣的领域：生物技术、工程学和计算机科学。在这种情况下，根本的临床问题不是仅仅通过采用传统的基于生物学的方法来解决的。然而，该项目的目标是开发一种多方面的方法，能够从生物、工程和计算机科学的角度解决临床问题。如果成功，这项工作将促进建立新药开发管道，可以进一步扩展到广泛的病理场景。与EPSRC的战略和研究领域保持一致该项目专注于使用人类衍生样本的人类临床应用。这一事实限制了在科学研究中使用动物，并遵守EPSRC通过的指导方针。博士项目的一致部分将涉及用于药物开发应用的机器学习(ML)和人工智能(AI)应用的开发。这项工作与EPSRC目前支持ML/AI方法改善医疗保健的兴趣一致。没有合作者。

Brief description of the context of the research including potential impactsDrug discovery is known to be a long and expensive process with very low success outcomes. The whole process usually develops over three phases: i) synthesis of drug molecules, ii) preclinical animal trials and iii) human clinical trials. Preclinical animal trials have shown crucial limitations in the past years, as caused by the wide interspecies gaps between animals and humans. This often translates into wide unreliability in animal trials' outcomes, with huge losses in pharmaceutical research and development programs.The use of human-derived in vitro/ex vivo models has opened new horizons in this field, by preserving patients-specific features and reproducing physiological and pathological scenarios more reliably. The enhanced accessibility to biological and clinical data has led to the development of more sophisticated computational models and bioinformatics applications. Machine learning models have shown great potential in therapeutic applications as they can provide novel predictive insights into drug efficacy by processing a vast number of heterogeneous datasets. The combination of novel coupled in vitro-in silico models has been described to be the crucial direction to follow to improve current drug development pipelines as expressed by relevant regulatory authorities (e.g., FDA, EMA). The urgency in establishing novel therapeutic strategies is evident in the field of liver disease, encompassing a wide range of pathologies with different aetiological factors (e.g., viral infections, alcohol abuse, extensive high-fat diet, etc.). Orthotopic liver transplantation is still considered the gold standard treatment for end-stage liver disease. However, this practice is still affected by crucial negative aspects namely highlighted in the necessity of long-term immunosuppressive treatments and the current shortage of organs suitable for transplantation. This situation is further exacerbated by the dramatic paucity of therapeutic alternatives in several pathological scenarios (e.g., nonalcoholic steatohepatitis - NASH, and biliary atresia), posing a global challenge to identify and develop novel effective treatments.Aims and objectivesThe doctoral project concerns the development of an engineered process aimed at the derivation and testing of multiple reliable in vitro models (e.g., organoids, precision-cut tissue slices) from patients' surgical samples. These models will be further used to assess the efficacy of novel drug candidates compared to standard-of-care treatment, by employing computational models and informatics resources. Novelty of the research methodologyThe project aims to establish a multidisciplinary approach to tackle current shortcomings in drug discovery processes by adopting a multifaceted perspective. The activity encompasses three major areas of interest: biotechnology, engineering, and computer science. In this setting, the underlying clinical problem is not tackled by solely employing a conventional biology-based approach. However, the project aims to develop a multifaceted approach able to solve the clinical problem from a biological, engineering, and computer science perspective. If successful, this work will promote the establishment of novel drug development pipelines that can be further expanded to a wide array of pathological scenarios.Alignment to EPSRC's strategies and research areasThe project focuses on human clinical applications with the employment of human-derived samples. This fact limits the employment of animals in scientific research and complies with guidelines adopted by EPSRC. A consistent part of the doctoral project will concern the development of machine learning (ML) and artificial intelligence (AI) applications for drug development application. This work aligns with EPSRC's current interest in supporting ML/AI approaches for improving healthcare.No collaborators.