Jonassen group

Pattern discovery in molecular biology data

Our research is focused on development and application of computational methods and tools for the analysis of molecular biology data. Our emphasis is on methods for the discovery of recurring patterns and unexpected correlations that may be of biological significance.

We are developing algorithmic tools for the automatic discovery of patterns (motifs) in protein sequences and structures. These are computationally very challenging problems.Therefore we focus on developing algorithms able to deal with large data sets efficiently exploring the space of possible patterns to identify the most significant motifs. This has so far resulted in the tools Pratt and SPratt for sequence and structure pattern discovery, respectively. We are currently taking the structure pattern idea further into structure model evaluation and structure prediction area.

Microarrays constitute a powerful technique for measuring gene expression levels. We are developing methods and software solutions for data representation including exchange of data and analyses allowing for accurate and machine-readable representations. This is of critical importance since both data and the analysis results are meaningful only in the context of a full description of how they were obtained. Furthermore we are working on developing and employing methods for clustering, classification, feature selection, as well as on methods for relating microarray data to sequence, network and annotation data. We see it as critically important to take part in collaborations with medical and biological research groups applying microarrays in joint projects.

While ESTs represent a valuable source of information about expressed genes and isoforms, individual ESTs are unreliable. Clustering and comparison with genomic information is therefore neccessary. We work on development of new methods for EST clustering, prediction of splice forms, and design of oligos to be used for detection using microarrays.

Future projects and goals:

• Take part in collaborative projects utilising microarray and other high throughput technologies to utilise bioinformatics and statistical methods for data analysis.
• Develop further methods for handling, processing, and analysing functional genomics data.
• Work on methods for the analysis, comparison, and prediction of protein structure.
• Develop fully automated methods for using EST data to generate gene indices, predict alternative splicing, and study implications on protein and regulatory level.

Selected publications:

• Bø, T. H., Dysvik, B., Jonassen, I. (2004) LSimpute: accurate estimation of missing values in microarray data with least squares methods. Nucl. Acids Res. 32:e34
• Malde, K., Coward, E., Jonassen, I. (2003) Fast Sequence Clustering Using A Suffix Array Algorithm. Bioinformatics 19:1221-1226
• Jonassen, I., Eidhammer, E., Taylor, W.R. (1999) Discovery of Local Packing Motifs in Protein Structures. Proteins: Struct., Funct., and Genet. 34:206-219
• Brazma, A., Jonassen, I., Vilo, J, Ukkonen, E. (1998) Prediction of Regulatory Elements in Silico on a Genomic Scale. Genome Research 8:1202-1215
• Jonassen, I., Collins, J. F., Higgins, D. G. (1995) Finding flexible patterns in unaligned protein sequences. Protein Science 4:1587-1595.