òòò½Íø

The Breitling group is interested in developing computational approaches to understand and engineer complex biological systems. A major focus of the work is directed towards empowering the predictive design of microbial cell factories to produce high-value chemical compounds. The group arrived at A*STAR in January 2025 and is currently recruiting new team members.

Our work is contributing to all stages of the Design–Build–Test–Learn cycle of synthetic biology. Some areas of special interest are the following:

Genome Mining and Natural Product Discovery: The Breitling group has pioneered antiSMASH, the industry-standard software for identifying and annotating the biosynthetic capabilities in microbial genome sequences, and continues to develop novel tools for exploiting this rich source of powerful biochemical machinery for applications in synthetic biology.

(Bio)Retrosynthesis and Pathway Design: Leveraging advances in AI to facilitate the identification of plausible biosynthetic routes to new compounds of interest and drive the design of novel enzymes to catalyse the required new-to-nature reactions efficiently.

Dynamic Systems Modelling: Using a wide range of modelling techniques, from resource-constrained genome-scale modelling approaches to differential equation systems, to predict the complex behaviour of cell signalling and cellular metabolism, e.g., to identify strategies for microbial strain optimisation in biotechnology. A particular focus is the explicit incorporation of uncertain, incomplete or semi-quantitative data in the modelling process and the integration of AI methods for improved parameter estimation for ensemble modelling.

Metabolomics: Developing new tools for the analysis of quantitative metabolomics data generated by high-resolution mass spectrometry, to understand cellular physiology and identify new biosynthetic capabilities in natural and genetically engineered organisms. This includes AI-driven approaches to automating the annotation of large collections on metabolite profiles for the matching of valuable metabolites to genomically encoded biosynthetic machinery.

Transcriptomics: Creating statistical techniques for gene expression analysis (RankProducts, iterative GroupAnalysis, VectorAnalysis), which are used in laboratories worldwide, and combining these with machine-learning strategies to identify diagnostic signatures that can, e.g., inform the debugging of engineered microorganisms created by synthetic biology approaches.

Many of the tools developed in the Breitling group are of broader interest for biomedical research applications, and the group is working closely with experimentalists interested in model organisms of various levels of complexity, ranging from bacteria, like the antibiotic producer Streptomyces coelicolor, to single-cell eukaryotic models such as the protozoan parasite Trypanosoma brucei, and finally to mammals, including mice and humans.