Bioinformatics and Big Data
Professor Inge Jonassen and part of his team from the CBU (Computational Bioinformatics Unit, Department of Informatics, UiB) are working on the development and application of bioinformatics methods for analysing data descending from high-throughput measurement technologies applied to cancer samples.
The Jonassen group is working on development and application of bioinformatics methods contributing to the understanding of tumors and their environments, aiming to aid in selecting appropriate treatments and predict outcome. A focus in earlier years has been on in silico deconvolution of expression data. New activity has been initiated towards system medicine approaches targeting leukemia and development of methods to exploit the novel Hyperion technology to the study of tumor-microenvironment interactions in solid cancers.
Jonassen is leading a project funded by ERAPerMed, including as partners Professor Gjertsen from CCBIO in addition to groups from Germany, the Netherlands and Canada. A postdoc has been recruited to work on this project in Jonassen’s group. The project includes data generation on single cell and bulk samples on genomic, transcriptomic and proteomic levels, systems biology modeling, and machine learning aimed at predicting outcome and aid selection of treatment for individual patients, use of a set of different experimental model systems and pilot clinical trials.
At the end of 2019, a new postdoc position within the Jonassen group was announced – to work on development and use of methods to exploit the novel Hyperion imaging technology to the study of tumor-microenvironment interactions.
Relevant to Jonassen’s work in CCBIO, he published (in bioRxiv) in 2019 a study of gene expression in Parkinson’s disease where it was shown that changes in cell composition confound expression signatures so far found to be associated with the disease. This was found by using a deconvolution approach utilizing previously known cell type signatures – an approach complementary to that earlier developed by the Jonassen group in context of CCBIO. This provides an improved basis for the work also within CCBIO in the coming period.
Plans for the future
The new EraPerMed project is tightly linked with CCBIO and will be an important focus for the Jonassen group in the coming three years. In addition, the group plans new efforts utilizing the Hyperion imaging platform, potentially together with single cell omics approaches to improve the understanding of tumormicroenvironment interactions.
Current challenges in the field
The Jonassen group aims to develop and use mathematical models that capture and predict effects of drugs targeting signaling molecules. Through the new project (above), they have established collaborations with groups having a strong track record in this area. In order to use such models to aid in selecting therapies for individual patients, they aim to utilize machine learning methods. One challenge is the relatively small size of training data that will be available for such approaches. The group’s approach will be to summarize the data and model predictions using a small number of parameters enabling learning from smaller training sets. A more technical challenge is the increasing focus from research funding agencies on data management plans and FAIR data sharing. This requires bioinformatics support, but also systematic efforts from those collecting samples and generating data in order to capture and describe in standardized ways meta-data allowing data reuse.