Molecular investigation of drug synergy in cancer

癌症治疗理论的一个主要挑战是预测每个患者对抗癌药物的实际临床反应。除了预测患者是否会对特定药物产生耐药性或敏感外，预测哪些药物组合可能对特定患者协同作用的能力也至关重要，即个性化药物方法。为此，我们创建了一个专家系统生物信息学途径，该途径模拟人类专家的决策能力，利用689个细胞系中139种药物的药物IC-50值，包括基因突变，基因调控和拷贝数数据，涵盖多种癌症。该系统已通过“重新发现”分子特征与药物敏感性之间的许多已知关联得到验证（提交的论文，2016）。它还能够“重新发现”协同作用的已知药物组合（例如BRAF和MEK抑制剂用于黑色素瘤）;综合起来，这证明了我们设计的新型生物信息学方法的成功。该项目将验证灵敏度和协同作用的新预测，并阐明协同关系的机制。本项目的目的是研究预测的协同药物组合的有效性和阐明机制。敏感性预测和特别感兴趣的药物组合之间的协同作用的预测将在现有的癌细胞系中进行验证，使用一组细胞系，其基因组和转录组分析可用于但最初未用于生物信息学预测。将进一步研究验证的协同作用，并阐明协同作用背后的机制。我们的生物信息学管道识别基因突变，表达值或拷贝数，这些基因突变，表达值或拷贝数在预测药物之间的协同作用方面可能至关重要（例如导致这些药物敏感性的常见突变）。这些预测标记物将用于鉴定靶分子以研究协同机制。然后将通过过表达、沉默或药理学调节感兴趣的靶标来探索机制。细胞死亡/衰老的机制和细胞分子应答也将使用各种适当的测定、免疫印迹、qPCR、强有力的药物组合候选物将被用于原代癌细胞培养的研究，如果时间允许的话，将用于体内模型的研究。如果成功，我们的项目将验证一种新的策略，用于鉴定临床有用的药物组合，以及验证特定的药物组合与临床潜力

A major challenge for cancer treatment rationales is predicting the actual clinical response to anti-cancer drugs for each individual patient. As well as predicting if a patient will be resistance or sensitive to a particular drug, the ability to predict which combinations of drugs may work synergistically for a particular patient is also critical i.e. the personalised medicine approach. To this end we created an expert system bioinformatic pathway that emulates the decision-making ability of a human expert, utilising drug IC-50 values for 139 drugs across 689 cell lines with gene mutation, gene regulation and copy number data, covering a multitude of cancers. This system has been validated by the "re-discovery" of many known associations between molecular signatures and drug sensitivity (paper submitted, 2016). It was also able to "re-discover" known drug combinations that work in synergy (e.g. BRAF and MEK inhibitors for melanoma); taken together this proves the success of the novel bioinformatics approach we have designed. This project would verify novel predictions of sensitivity and synergy as well as elucidating the mechanisms of synergetic relationships. The objective of this project is to investigate validity and elucidate mechanisms of predicted synergistic drug combinations. Sensitivity predictions and predictions of synergism, between drug combinations of particular interest, will be in validated in existing cancer cell lines, using a panel of cell lines for which genomic and transcriptomic profiling is available for but were not originally used in bioinformatics predictions. Validated synergism will be further investigated and the mechanism behind the synergy elucidated. Our bioinformatic pipeline above identifies gene mutations, expression values or copy numbers that are potentially crucial in predicting the synergy between drugs (e.g. common mutations which lead to sensitivity of these drugs). These predictive markers will be used to identify target molecules for investigations into synergistic mechanisms. The mechanism will then be probed by over expression, silencing, or pharmacological modulation of targets of interest. The mechanism of cell death/senescence and the cells molecular response will also be probed using various appropriate assays, immunoblotting, qPCR, etc.Strong drug combination candidates will be taken forward for investigation in primary cancer cell cultures and if time allows to in vivo models.If successful our project will validate a novel strategy for identifying clinically useful drug combinations as well as validating specific drug combinations with clinical potential.