The first part of the talk reviews Bims: a system for Bayesian machine learning. The theory of the
  system is explained by providing details on (a) the syntax of distributional logic programs for
  expressing prior knowledge and (b) the integrative mechanics with which Markov chain Monte Carlo
  inference is employed. Comparative results to support vector machines and neural networks are shown
  for a ligand biding dataset. The second part of the talk spans a number of projects within cancer
  data analytics. These include analyses of microscopy and proteomics datasets. Constructing Bayesian
  networks for genomic data from a number of cancer types is discussed along with additional layers
  of information we have pioneered as to elucidate network relations to biologists. Aspects of modern
  logic programming that have been crucial to working with large biological datasets are highlighted
  with contributed libraries mentioned in context. The talk concludes with observations on the role
  of statistically aware symbolic approaches to modern AI and particularly within computational
  biology. Knowledge is proffered as a key concept and tool in the search for explainable,
  accountable, open and shareable AI.