Screening protein sequences to discover novel protein functions using informatics target selection and ultrahigh-throughput droplet microfluidics

博士项目战略主题：生物科学可再生资源和清洁生长细菌群落作为生物催化剂的重要来源提供了潜力，然而它们中的许多是不可培养的，因此潜在的新颖和有用的生物催化剂仍未被发现。此外，由于新的生物催化剂的预期活性不一定具有相似的序列，序列分析不能预测许多新的命中。利用诱饵底物的微流控微滴室进行功能筛选可以鉴定这些新蛋白。液滴微流体是一种将单个细胞隔离在水-油乳液中并在微通道中流动的技术。文献中已经创建了许多模块来扩展微流体装置的工作流程，包括流动聚焦装置、皮注射装置和吸光度激活的细胞分选装置。液滴微流体的好处包括它可以促进超高通量实验（可以根据每小时超过100万个液滴的吸光度读数进行分类），这需要挑选出罕见的变异，它需要最少的试剂量，液滴可以独立操作，作为单独的反应容器。重点将是利用微滴微流体技术，利用宏基因组数据来发现新的酶。这将是计算和实验方法之间的协同作用，可以进行交叉比较，并弥合湿实验室/硅的鸿沟。当前的微流控模块将被迭代，以允许更大的微流控工作流程自动化。机器学习方法将用于更好地描述获得的大量实验数据。通过在皮升反应体积中进行快速实验，计算方法将通过湿/干工作的迭代循环得到改进，并且基于阵列的基因合成将被利用来允许访问所有预测，以发现新的生物催化剂。

PhD project strategic theme: Biosciences for renewable resources and clean growthBacterial communities provide potential as a great source for biocatalysts, however many of them are unculturable and therefore potentially novel and useful biocatalysts remain undetected. Additionally, because desired activities of novel biocatalysts do not necessarily share similar sequences, sequence analysis can not predict many novel hits. Functional screening using microfluidic picodroplet compartments using bait substrates can identify these novel proteins.Droplet microfluidics is a technique in which single cells are compartmentalised in water oil emulsions which flow in microchannels. A number of modules have been created in the literature to expand the workflow of the microfluidic devices including flow focusing devices, picoinjection devices, and absorbance activated cell sorting devices. The benefits of droplet microfluidics include that it can facilitate ultrahigh-throughput experiments (can sort based on absorbance readout of greater than 1 million droplets per hour) which is needed to pick out rare variants, it requires minimum amount of reagents, and droplets can be manipulated independently, serving as individual reaction vessels.The focus will be on exploiting metagenomic data to discover novel enzymes by using droplet microfluidics. It will be a synergy between computation and experimental approaches enable cross-comparison and bridge the wet lab/in silico divide. Current microfluidic modules will be iterated on to allow greater automation of the microfluidic workflow process. Machine learning methods will be used to better characterise the large amount of experimental data gained. By performing rapid experiments in picolitre reaction volumes the computation methods will be refined by iterative cycles of wet/dry work, and array based gene synthesis will be exploited to allow access to all predications, in order to discover novel biocatalysts.