Project Summary/Abstract Asthma is one of the most common chronic respiratory diseases worldwide. Microbial dysbiosis in the gut and lungs has increasingly been associated with the incidence and severity of asthma, indicating the potential of the microbiome to be a determinant factor in asthma pathogenesis. However, as the most likely connection between the gut and lungs, the role of the blood microbiome in the “gut–lung axis” is still unclear for asthma pathogenesis due to the lack of cost-effective and high-throughput sequencing methods. Indeed, it is either impossible, or prohibitively expensive, for conventional sequencing methods to handle microbial DNA samples that are in trace amounts, heavily degraded, or dominated by host DNA, e.g., in human blood. We hypothesize that the presence of microorganisms in the blood is related to the risk of asthma occurrence, and these microbial blood biomarkers can be captured by a reduced metagenomic sequencing method for diagnosis or even early detection of asthma with the help of a deep learning framework. In this application, a strain-resolved computational pipeline for the reduced metagenomic sequencing will be developed to profile the blood microbiome in the “Vitamin D Antenatal Asthma Reduction Trial” --- an ongoing randomized, double- blind, placebo-controlled clinical trial of 881 pregnant women with both questionnaires and maternal, cord, and child blood samples available. Meanwhile, a deep-learning framework will be developed to optimize the accuracy of diagnosis and prediction models for asthma using blood microbiome data. Finally, with the aid of the new computational pipeline and deep-learning framework, the role of the blood microbiome in the gut– lung axis and asthma pathogenesis will be investigated. Dr. Sun’s trainings in molecular biology, bioinformatics and metagenomics have prepared him well for this proposed research. However, understanding the molecular basis connecting asthma through the analysis of blood microbiome data is a challenging task that requires further training in specific areas. Dr. Sun will leverage the excellent intellectual environment of Harvard Medical School (HMS) and its teaching hospital Brigham and Women’s Hospital (BWH). He will have access to extensive computational resources at BWH and HMS. Through formal coursework and workshops, and with the help of a strong mentoring team and a world-class advisory committee with complementary expertise, Dr. Sun will immerse himself in a training program focusing on advanced programming, statistical modeling, deep learning, respiratory pathophysiology, and clinical translation. Dr. Sun will meet with his two mentors and advisory committee members on a regular or needed basis to present his progress and get prompt feedback and advice. Altogether, Dr. Sun’s training and research plan will enable him to expand his current skill set to include the ability to address the challenges of low microbial biomass sequencing in blood s...