RMISTCR - Rapidly mining the immune system for rare therapeutic T-Cell Receptors to treat solid tumour cancers

* * 需要 ** 根据英国癌症研究中心的数据，英国每年诊断出约367，000例新癌症病例。在英国，每年有16.5万人死于癌症。目前迫切需要"新的治疗实体瘤癌症的治疗方法.. * *挑战 ** 基于T细胞受体的细胞疗法是治疗无法治愈的晚期实体瘤的有希望的治疗方法。它们由称为T细胞的免疫细胞组成，来自患者，然后用癌症靶向T细胞受体（TCR）进行基因重编程，然后繁殖，并重新引入患者体内以摧毁癌细胞。不幸的是，找到这些治疗性TCR是大海捞针的问题，也是使用实验室现有技术的主要挑战，例如哺乳动物展示，其平均发现时间约为6年，每次活检的成本为1000万英镑。迫切需要计算方法来自动化和简化这一过程。创新 ** 我们打算通过使用先进的人工智能（神经网络和深度学习算法架构）和新型高通量库对库湿实验室筛选技术（慢病毒展示）来消除这一障碍，以识别罕见的癌症靶向TCR，这将使我们的平台能够快速筛选患者的数十亿TCR，以便通过计算识别可以靶向和破坏癌细胞的罕见TCR。AI从我们在实验室使用慢病毒展示分析的TCR和癌细胞之间的数十亿次相互作用中学习。影响 ** 与目前可用的展示分析不同，它专注于一次分析一个TCR，我们创新的湿实验室+AI方法可以同时进行快速TCR筛查，有助于在约4周内全面筛查数十万个TCR，（约£ 3.5k/活检），大大增加了识别罕见癌症特异性TCR的机会。这是发展生命的基本垫脚石-为各种无法治疗的实体瘤保存细胞疗法，并帮助全球数百万患者与癌症作斗争。该技术将加快英国药物发现，加速开发用于以前无法治疗的实体瘤的新的挽救生命的癌症治疗方法，并最终降低NHS的治疗成本。生物医学催化剂项目将验证我们的方法，加强我们的人工智能训练数据，使我们能够使用真实的血液样本证明临床有效性。从商业角度来看，项目产出的结果将与英国世界领先的TCR治疗开发商Immunocore建立合作关系。

**Need**According to Cancer Research UK, ~367,000 new cancer cases are diagnosed every year in the UK. Cancer kills 165,000 people each year in the UK. There is the urgent need for "new curative treatments for solid tumour cancer... cheaply and effectively" (NHS-Long Term Plan).**Challenge**T-cell-receptor-based cell therapies are promising curative treatments for otherwise untreatable advanced solid tumours. Consisting of immune cells known as T-cells, they are derived from patients, then genetically reprogrammed with a cancer-targeting T-cell receptor (TCR), before being multiplied, and reintroduced into patients to destroy cancer cells.Unfortunately finding these curative TCRs is a needle in a haystack problem and a major challenge using existing technologies in the lab such as mammalian display, which has an average discovery time of ~6 years and costs \\\>£10million/biopsy. There is an urgent need for computational methods to automate and streamline the process.**Innovation**We intend to remove this roadblock to identifying rare, cancer-targeting TCRs through the use of advanced AI (neural networks and deep learning algorithm architecture) and a novel high-throughput library-on-library wet lab screening technique (lentiviral display).This will enable our platform to rapidly screen billions of TCRs from patients in order to computationally identify rare TCRs that can target and destroy cancer cells. The AI learns from the billions of interactions between TCRs and cancer cells that we analyse at our labs using lentiviral display.**Impact**Unlike currently available display assays, which focus on the analysis of one TCR at the time, our innovative wet-lab+AI approach enables rapid TCR screening to run concurrently, facilitating the comprehensive screening of hundreds of thousands TCR in ~4 weeks, (~£3.5k/biopsy), dramatically increasing the chances of identifying rare cancer specific TCRs.This is a fundamental stepping stone to developing life-saving cell therapies for a wide range of untreatable solid tumours and helping millions of patients worldwide in their fight against cancer.The technology will speed-up UK drug discovery, accelerate the development of new life-saving cancer treatments for previously untreatable solid tumours, and ultimately reduce treatment costs to the NHS.The Biomedical Catalyst project will validate our approach, strengthen our AI- training data and enable us to demonstrate clinical validity using real blood samples. Whilst from a commercial perspective, the results of the project outputs will unlock a partnership with Immunocore, UK world-leading TCR therapy developer.