Data has become critical within the research environment with many academic and industry teams producing data on healthy and disease-based populations across many sample types. With the growth in Omics technologies, data libraries are expanding rapidly, and in addition, endpoints on clinical trials are increasing and becoming more digitised. The methodology for analysing the data generated is also expanding with the use of knowledge graphs, ML/AI methods and algorithms to assess trends and make predictions using the data, creating the potential to revolutionise academic research and drug discovery.Fulfilling the potential of these data is a challenge; how do we share these data externally, what are the best platforms and for what data science applications? At AstraZeneca, through our Open Innovation programme we offer researchers access to high quality preclinical data. These data can be used for repurposing to enable researchers to derive wider insights into biological mechanisms and outcomes, creating broad benefit through collaboration with the scientific community.We currently offer access to our Oncology combinations dataset, where 109 molecules were screened in a 7x7 concentration matrix over a 1000-fold range across the IC50 in 755 cell lines. Cell viability data and associated downstream analysis data are available. These data have been accessed by multiple groups, including in a study led by Dr Jing Tang who used the data to develop Drug Comb (https://ai.drugcomb.org/). This, along with data from other internal studies, has been impactful across several external academic groups.As data is generated across global research groups in AstraZeneca, we see real power in sharing this to benefit researchers in developing methods and gaining further understanding of disease.
Zheng Shuyu, Aldahdooh Jehad, Shadbahr Tolou, Wang Yinyin, Aldahdooh Dalal, Bao Jie, Wang Wenyu, Tang Jing. DrugComb update: a more comprehensive drug sensitivity data repository and analysis portal. Nucleic Acids Research. Vol. 49(W1)2021. Oxford University Press (OUP). [Cross Ref]