Computational Biology and Bioinformatics

The main interests of the CBB group are comparative genomics, gene regulation, and cancer progression. Our goal is to develop biological relevant methods, by using advanced probabilistic modeling and algorithm design,

Comparative Genomics

In a genome, the genes evolve through nucleotide substitutions. The evolution of the genome is shaped by a multitude of evolutionary events acting at different organizational levels. Larger genome segments are affected by processes such as duplication, lateral transfer (where a segment of an organisms genome is transfered to the genome of another organism), inversion, transposition, deletion and insertion. We have developed an integrated probabilistic model for gene duplications, gene losses, and sequence evolution together with analysis algorithms as well as tools based on this model. Lateral transfer is another evolutionary event that has gained a lot of our attention. We have developed parsimony algorithms for lateral transfer identification and are currently working on a probabilistic model that contains duplications, losses, and lateral transfers.

Gene regulation

Regulatory elements, or transcription factor bindings sites, appear in clusters typically upstream of coding regions of the genes they take part in regulating. Identification of regulatory elements is an important step towards revealing regulatory networks. We have developt methods for genome wide identification of modules of regulatory elements as well as tree-based methods for identification of individual regulatory elements.

Cancer informatics

Chromosomal aberrations in solid tumors appears in complex patterns. It is important to understand: how these patterns develop, the dynamics of the process, the temporal or even causal order between aberrations, and the involved pathways. We are interested in mathematical models, algorithms and automated tools that enable derivation of network models from various cancer data sets. A network consist of vertices representing chromosomal aberrations and directed edges, between such vertices, representing temporal and causal relations. We have developed, and applied, algorithms that can derive network models for cytogentic data and are in the process of extending these results to other data forms.