ReportTitle: Understanding seabet io from Pathogen Genetic Data
Report time:2019Year12month16Sunday afternoon4:00
Reporting location:School Main Computer Building206Conference Room
Reporter: Xu YuanweiDoctor
Abstract: Whole genome sequencing (WGS) has been routinely implemented for infectious disease outbreak investigations due to the decreasing cost of sequencing. Understanding person-to-person seabet io events and seabet io patterns is essential to public health in guiding their strategies for outbreak control. In the case of TB, previous studies have shown that WGS of patient isolates showed higher agreement with contact investigations than previous biomarkers. WGS for outbreak reconstruction typically involves the following steps: isolate collection, sequence analysis, seabet io clusters identification and seabet io inference. In this presentation I will focus in particular on our work of Bayesian seabet io reconstruction, a two-stage approach of first constructing a phylogenetic tree from the sequences, then Bayesian inference of seabet io trees given the phylogenetic tree. I will show how the inference can be done efficiently when faced with many seabet io clusters. As we gather more data from patients, how to integrate different sources of data including genetic, epidemiological and patient-level data is becoming an important area of research. I will show our attempt to use machine learning to predict credible transmitters from covariate data such as demographic and social-economic variables.
