13 May 2024 - 13 May 2024
1:00PM - 4:00PM
Room MCS2068, Mathematical Sciences & Computer Science Building, Durham University, Upper Mountjoy Campus, Stockton Road, Durham DH1 3LE
Free
Biostatistics Unit Seminar with speaker Dr Heather Cordell from Newcastle University on the topic of the bayesian network approach.
Maths Building in the distance with grass in the foreground
Seminar Title - A Bayesian network approach incorporating imputation of missing data allows exploration of complex causal biological relationships between variables associated with type 2 diabetes.
Abstract: Bayesian networks can be used to identify possible causal relationships between variables based on their conditional dependencies and independencies, which can be useful in complex biological scenarios with many measured variables. In practice, many large scale biological data sets include observations with missing data. We previously developed an imputation approach that capitalises on the initial inferred network structure to allow the recovery of missing data, providing subsequent improvement to the final inferred network or networks. We demonstrate the performance of our proposed approach on simulated data and apply it to data from a large pan European study of type 2 diabetes (T2D) in which not a single individual had complete data for all variables. Our results confirm many previous findings in T2D (such as the relationship with liver fat) and highlight potential newavenues for investigation.
This will be a hybrid event via zoom click here to join on the day.