Search for: All records

Abstract Background Applying directed acyclic graph (DAG) models to proteogenomic data has been shown effective for detecting causal biomarkers of complex diseases. However, there remain unsolved challenges in DAG learning to jointly model binary clinical outcome variables and continuous biomarker measurements. Results In this paper, we propose a new tool, DAGBagM, to learn DAGs with both continuous and binary nodes. By using appropriate models, DAGBagM allows for either continuous or binary nodes to be parent or child nodes. It employs a bootstrap aggregating strategy to reduce false positives in edge inference. At the same time, the aggregation procedure provides a flexible framework to robustly incorporate prior information on edges. Conclusions Through extensive simulation experiments, we demonstrate that DAGBagM has superior performance compared to alternative strategies for modeling mixed types of nodes. In addition, DAGBagM is computationally more efficient than two competing methods. When applying DAGBagM to proteogenomic datasets from ovarian cancer studies, we identify potential protein biomarkers for platinum refractory/resistant response in ovarian cancer. DAGBagM is made available as a github repository at https://github.com/jie108/dagbagM . 

Recent evidence suggests an association between endometrial cancer and the understudied bacterial species Porphyromonas somerae . This association was demonstrated in previous work that indicated a significantly enriched abundance of P. somerae in the uterine microbiome of endometrial cancer patients. Given the known associations of the Porphyromonas genus and oral cancer, we hypothesized that P. somerae may play a similar pathogenic role in endometrial cancer via intracellular activity. Before testing our hypothesis, we first characterized P. somerae biology, as current background data is limited. These novel characterizations include growth curves in liquid medium and susceptibility tests to antibiotics. We tested our hypothesis by examining growth changes in response to 17β-estradiol, a known risk factor for endometrial cancer, followed by metabolomic profiling in the presence and absence of 17β-estradiol. We found that P. somerae exhibits increased growth in the presence of 17β-estradiol of various concentrations. However, we did not find significant changes in metabolite levels in response to 17β-estradiol. To study direct host-microbe interactions, we used in vitro invasion assays under hypoxic conditions and found evidence for intracellular invasion of P. somerae in endometrial adenocarcinoma cells. We also examined these interactions in the presence of 17β-estradiol but did not observe changes in invasion frequency. Invasion was shown using three lines of evidence including visualization via differential staining and brightfield microscopy, increased frequency of bacterial recovery after co-culturing, and in silico methods to detail relevant genomic and transcriptomic components. These results underscore potential intracellular phenotypes of P. somerae within the uterine microbiome. Furthermore, these results raise new questions pertaining to the role of P. somerae in the progression of endometrial cancer.