< Back to previous page

Project

Probabilistic Logics for Biological Network Inference.

ProbLog is a probabilistic extention of the programming language Prolog, which has been developed at the K.U.leuven. It can be used to calculate the succes probability of a query, based on probabilistic facts and deterministic Prolog rules. The main motivation for the creation of this language was the emergence of large probabilistic networks in the field of Bio-informatics. These networks are the result of high-throughput experiments that generate a large number of results that have a probabilistic nature. The field of Bio-informatics uses techniques of artificial intelligence, in order to infer knowledge from this data. For this task, several specific algorithms have been created in the bioinformatics literature, which are often based on specific
 mechanisms for dealingwith the probabilistic nature of the data. This is a very hard task andthe creation of these algorithms takes a lot of time. It is the goal ofthis doctorate to investigate whether a general purposes probabilistic language such as ProbLog can be used for inference in such biological networks. When using ProbLog, one can directly rely on a probabilistic semantics as well as efficient inference mechanisms. Furthermore, by modifying the rules it should be possible to emulate several of the existing algorithms.
 

Date:1 Jan 2013 →  31 Jul 2015
Keywords:Logics for Biological, Bioinformatics, Artificial Intelligence, probabilistic programming
Project type:PhD project