Zheng Jie’s team uses genetic big data to build a causal network between proteins and human diseases

Using Mendel’s randomization method, this study constructed a complete causal network between 1002 proteins and 225 complex human diseases. By comparing this causal network with the existing drug randomized double-blind experiment, this study confirmed that the inference of causal relationship between protein and human disease can effectively improve the success rate of drug RCT, and deeply revealed the importance of proteomic genetic research methods in drug target early screening. As the direct target of most drugs, < / P > < p > proteins play a central role in drug research and development. Most of the drugs used to treat human diseases will regulate people’s traits by changing the level of protein in the human body, so as to reduce the risk of human diseases. Therefore, constructing the causal relationship between proteins and human diseases is one of the core issues in drug research and development. < / P > < p > randomized controlled trials are the golden rule to construct the relationship between drugs and diseases. But its cost is expensive, time-consuming and laborious, and the failure rate is still high. The development of genetic big data and Mendel’s randomization method makes it possible to search for the causal relationship between proteome and human diseases on a large scale. This kind of research plays an important role in finding suitable drug targets, increasing the success rate of randomized controlled trials and reducing the cost of drug development. However, such genetic research is still in “infancy” [2,3], large-scale experiments are needed to explore its feasibility, and relevant research standards need to be formulated to regulate future research. < / P > < p > in the past few decades, the investment in drug R & D has been increasing, but the success rate of drug R & D has been declining. The reason is that more than 50% of the drugs are ineffective and 25% of the drugs have excessive side effects; however, these drugs have been put into the expensive phase III and phase IV randomized controlled trials. Therefore, an effective screening system for early drug targets plays an important role in reducing the high cost of drugs. Human genetics is expected in the screening of drug targets because its research object is human itself. Since the advent of genome-wide association analysis in 2006, how to quantify the value of human genetics in drug research and development is still an open question. Recently, some research results in the field of genetics have made it possible to systematically search for drug targets. First, the in-depth exploration of proteomics genetics makes it no longer a dream to explore and study thousands of proteins at the same time; second, the opening of the use right of large sample biological databases such as UK Biobank and the establishment of genetic big data application platforms such as Mr base and LD hub [4, 5, 6] make it possible to study hundreds of human diseases at the same time. On the basis of these studies, Zheng Jie’s team systematically used Mendelian randomization to establish a causal network between 1002 proteins and 225 human diseases, and revealed significant causal relationships between 65 proteins and 52 human diseases. < p > < p > compared with the existing randomized controlled trials, it was found that the success rate of protein targets which were supported by Mendel’s randomized screening evidence was significantly improved. < / P > < p > 2) by screening the relationship between the drug and other diseases, we can effectively prompt the related side effects of the drug; < / P > < p > 3) by searching the existing drugs, we can explore the therapeutic effect of the existing drugs on other diseases, and find out the possibility of new use of the old drugs. < p > < p > 1) for the first time, it systematically expounded that linkage disequilibrium would cause deviation to 31.5% of Mendelian randomization results, and further emphasized the importance of CO localization and other follow-up methods in such studies, which provided reference for standardizing the research strategy of proteomics; < / P > < p > 2) in order to effectively increase the reliability of CO localization method in multi-point region, pair wi was proposed This method was used to detect the causal relationship between 23 previously undetected proteins and human diseases. In order to support open scientific research and help establish research standards, Zheng Jie’s team published the analysis scheme of this study and the analysis results of more than 220000 pairs of protein disease causal relationship. < / P > < p > in order to facilitate users to find out the relationship between drug targets and diseases at any time, the team also established a visual data platform epigraphdb. < / P > < p > figure: the diagram shows the causal relationship between 111 pairs of proteins and human diseases. It covers 8 categories of diseases, such as cardiovascular diseases and cancer. The labeled proteins such as PCSK9 have been put into practical treatment as drugs. To sum up, this achievement is the frontier research of genetic epidemiology, which is dominated by achievement transformation. It explores and standardizes the genetic research process of proteomics from the perspective of methodology, and lays a solid foundation for future interdisciplinary research. Due to the deep cooperation with pharmaceutical enterprises, the research also provides a new idea for the early R & D process of pharmaceutical enterprises, and takes an important step in the transformation of genetic achievements. < p > < p > from the results, this study for the first time systematically demonstrated that the genetic research method of proteomics can effectively improve the success rate of drug research and development, confirmed its importance in drug target screening, and revealed more than 100 new drug targets. In order to reduce the cost of drug research and development, and then reduce the price of drugs, provide effective methods and ways for human health. < p > < p > it is reported that among the authors of this paper, Assistant Professor Zheng Jie of MRC comprehensive epidemiology research center of Bristol University is the first author and corresponding author of the paper, and his co-authors are associate professor Robert A. Scott and Professor Tom gaunt of Bristol University. The current research direction of Zheng Jie’s team is the application of statistical genetic model and causal inference of complex diseases in pharmaceutical industry. Pigtail is the most delicious way to do, super simple, women often eat, beauty and beauty, delay aging