A Method For Oxidative Stress Evaluation In Human Tissues And Its Applications On Cancer Research

Oxidation is a normal and necessary process that occurs in the human body all the time,and the redox balance is also one of the important chemical balances in the human body.Many biochemical reactions in the human body produce reactive oxidative substances.For example,when the immune system kills pathogens,it will use respiratory burst to kill germs by producing a large amount of peroxide.However,when there is an excess of reactive oxidative substances and the body’s antioxidant system cannot provide the same reducing ability,too much reactive oxides will damage proteins,lipids,DNA,etc.,resulting in programmed cell death.When certain biological conditions are met,various diseases may also be induced,such as Crohn’s disease,Alzheimer’s disease,and cancer.Therefore,this paper takes oxidative stress as the research object,trying to estimate the level of oxidative stress and study its role in the cancer metabolic network.There are many studies on the relationship between oxidative stress and disease.The main method used is to design different control groups,measure certain biomolecules produced by the samples,and then perform statistical analysis.As the scale and availability of bio-omics data is greatly improved,this provides a new direction for studying biological problems-using high-throughput omics data to study biological problems.The current work on using omics data combined with algorithmic strategies to study oxidative stress is at a very early stage.Through our research,we can obtain a lot of valuable information from omics data,and provide new ideas and clues for the study of biological problems.An unavoidable problem in the study of oxidative stress is to measure or evaluate the severity of oxidative stress.Being able to accurately and reliably estimate the level of oxidative stress is of great significance for further research on the impact of oxidative stress on organisms.There is currently no publicly available tool for estimating oxidative stress from omics data,nor is there a publicly available dataset that includes both omics data and oxidative stress measurements.Therefore,this paper attempts to provide a model that can estimate oxidative stress based on transcriptome data to facilitate various downstream analyses,such as mean comparison,survival analysis,correlation analysis,and so on.Another important question in the study of oxidative stress is how oxidative stress drives the occurrence and development of diseases.Numerous studies have shown that environmental stress causes diseases generally through gene mutation to change some metabolic processes of cells,or directly affect some aspects of the metabolic process,and then lead to the occurrence of diseases.By exploring the network of oxidative stress and metabolic pathways and metabolic enzymes,we can discover how oxidative stress affects the entire metabolic network by affecting certain chemical reactions.In this paper,a graph neural network architecture based on variational autoencoder is used for network structure inference.Taking liver cancer data as an example,the role of oxidative stress in cancer cell metabolism is studied,which provides a new way for cancer mechanism research.ideas and plans.The main contributions of this paper are as follows:(1)A new quantitative estimation algorithm for oxidative stress is proposed,and the correctness of the method is verified on different datasets.By applying this model,multiple inferences related to oxidative stress were obtained,such as oxidative stress was positively correlated with the degree of cancer malignancy,and oxidative stress was an important factor in promoting the development of chronic inflammation to cancer..(2)Use graph neural network for directed graph structure inference,analyze the position and role of oxidative stress in the metabolic network of cancer cells,and then analyze and explain how oxidative stress affects biological metabolic pathways by affecting local biochemical reactions,thereby affecting the entire metabolism network,causing global abnormalities in cellular metabolism,leading to the development and progression of cancer.The experimental results show that our prediction results can provide reference for cancer mechanism research and help to discover new drugs or therapeutic targets.