Welcome to the Karakach Lab

The laboratory of integrative
multi-omics research at the Department of Pharmacology, Faculty of Medicine, Dalhousie University

500px-Dalhousie_University_Faculty_of_Medicine.svg.png
Researcher Performing Experiment

Our Research

The main objective of our research is to develop novel bioinformatics approaches that allow access to biological information embedded in the large amounts of data generated by bio-analytical technology platforms. The methods we develop push the limit in modelling low abundance signals that still exhibit important biological influence for key phenotypes such as paediatric malignancies. 

 

 

Our integrative approach to data analysis is driven by an emerging trend where biological data are analyzed simultaneously to answer specific research questions, leading to an accurate, comprehensive understanding of the correspondence between empirical observations, bioinformatics models and principal biological truths related, e.g., to oncogenesis, molecular sub-typing, metastasis, relapse and personalized chemotherapeutic response.

The Three Pillars

Keyboard and Mouse

Fundamental Computational Method Development

Error Modelled Gene Expression Analysis (EMOGEA), is a framework for  Functional Genomics data analysis that incorporates measurement errors in the analysis while introducing a special formulation for modeling time-course and/or ordinal measurement data.  By incorporating measurement errors, EMOGEA is specifically suited for measurements in which fluctuations in the low-level signals can lead to significant biological effects.  Examples include signaling mRNAs and non-coding RNAs (ncRNA) that are known to exhibit low levels of expression. We use this principle in our computational developments including our machine learning approaches for data visualization. 

Translational Bioinformatics

Lack of user-friendly software remains one of the biggest challenges in -omics research and is the limiting factor in access to the vast information inherent in these data. Our lab is developing a GUI-based tool referred to as the Bioinformatics Toolkit for CLinicians and EXperinentalists (BITCLEX). It is a user-friendly software that is comprised of a complete suite of bioinformatics pipelines used in multi-omics data analysis and visualization for non-data scientists. 

We also include a bioinformatics pipeline for analysis of High Throughput Sequencing (HTS) data employed in genotyping by sequencing, to allow clinicians to quickly visualize their results.

Functional Biology via Empirical Bio-Analytical Measurements

Molecular biology tools provide data that allow insight into important cellular processes and heterogeneities in their biology as it relates to their response to external perturbations such as disease, or chemotherapeutics. We design experiments that take full advantage of high throughput bioanalytical instruments to answer questions related to frailty in aging, pediatric cancers among others with the goal of understanding molecular mechanisms that can help us understand the etiology of the phenotype, their prognosis or to identify molecular pathways that we can target for developing personalized treatments.

Our lab approaches research in Pharmacogenomics in an integrated fashion with our activities divided into 3 interconnected pillars. These pillars are: fundamental bioinformatics, translational bioinformatics, and empirical molecular biology.  In fundamental bioinformatics, we develop computational tools to analyze and manage the large-scale -omics data that have come to define research in molecular biology. In the translational arm of our lab, we develop tools that can be used in a translational way where the methods from pillar one can be applied to model existing data e.g. from biorepositories. In addition, we develop GUI-based tools to allow experimentalists to analyze -omics data without a need for a detailed understanding of the statistical or computational methods. We test these tools to ensure that their applications are robust and provide guidance on their use. In the empirical molecular biology pillar, we employ analytical tools to validate our computational results while also designing experiments to test specific hypotheses.

Screen Shot 2021-10-05 at 10.34.39 PM.png