Marcin P. Joachimiak

Divisions

Environmental Genomics and Systems Biology

• Biosystems Data Science

Biography

Marcin Joachimiak is a staff researcher in the Environmental Genomics and Systems Biology Division at Lawrence Berkeley National Laboratory (Berkeley Lab). His current research is performed as part of the DOE Systems Biology Knowledgebase (KBase), the NIH Bridge2AI project, as well as a Berkeley Lab-funded project for predicting microbial growth conditions. He has twenty-five years of experience and a publishing record (H-index 30) in developing and evaluating machine learning algorithms, methods for functional genomics, bioinformatics data science, knowledge modeling, and computational systems design for computational biology. For the past 17 years he has been a member of large distributed microbial and biomedical projects with either a large computational component interfacing with experimentalists (DOE ESPP, DOE ENIGMA Science Focus Area) or which are computational systems centered around modeling and inferring knowledge (KBase, Monarch Initiative, NCATS Biomedical Data Translator, NIH Phenomics First, NIH Bridge2AI). He has led or currently leads team efforts in functional genomics analysis (ENIGMA, KBase, NCATS Translator), knowledge modeling (KBase, Bridge2AI), and machine learning algorithm development (ENIGMA, KBase, NCATS Translator).

Research Interests

Marcin Joachimiak is a computational biologist dedicated to the integration of microbial biology with machine learning and semantic technologies. At the forefront of my research interests is the development of machine learning techniques, including graph learning and biclustering, to unravel the complex relationships in microbiology and metagenome data. Additionally, by leveraging semantic technologies and ontologies, I aim to enhance machine learning, data science, and knowledge graph construction for microbes and microbiomes – a more nuanced and dynamic framework for studying microbial life. As an example, I am developing advanced models for growth prediction and culturing medium optimization using knowledge graphs and graph learning, including tailoring to the unique environment characteristics of different microbiomes. This research has the potential to improve our ability to characterize microbes and develop new utilitarian purposes such as in biomanufacturing, environmental remediation, and human health. Via synthesis of these computational methods and harmonized metagenome and microbiological data, new possibilities emerge in managing and manipulating microbes and microbiomes.