Machine learning illuminates how diet influences the evolution of yeast galactose metabolism

Harrison, Marie-Claire; Ubbelohde, Emily J.; LaBella, Abigail L.; Opulente, Dana A.; Wolters, John F.; Zhou, Xiaofan; Shen, Xing‐Xing; Groenewald, Marizeth; Hittinger, Chris Todd; Rokas, Antonis

doi:10.1101/2023.07.20.549758

Cited by 3 publications

(2 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This comprehensive dataset and analytical framework provide the opportunity to study how the observed genomic, metabolic, and environmental variation across the subphylum is associated with any complex trait of interest (68)(69)(70). To illustrate this potential, we examined the metabolic niche breadths of yeast pathogens of humans compared with those of their nonpathogenic close relatives (using a specific phylogenetic distance cutoff to standardize the clades) (Fig.…”

Section: Human Yeast Pathogens Include Both Carbon Generalists and Sp...mentioning

confidence: 99%

Genomic factors shape carbon and nitrogen metabolic niche breadth across Saccharomycotina yeasts

Opulente,

LaBella,

Harrison

et al. 2024

Science

Self Cite

View full text Add to dashboard Cite

Organisms exhibit extensive variation in ecological niche breadth, from very narrow (specialists) to very broad (generalists). Two general paradigms have been proposed to explain this variation: (i) trade-offs between performance efficiency and breadth and (ii) the joint influence of extrinsic (environmental) and intrinsic (genomic) factors. We assembled genomic, metabolic, and ecological data from nearly all known species of the ancient fungal subphylum Saccharomycotina (1154 yeast strains from 1051 species), grown in 24 different environmental conditions, to examine niche breadth evolution. We found that large differences in the breadth of carbon utilization traits between yeasts stem from intrinsic differences in genes encoding specific metabolic pathways, but we found limited evidence for trade-offs. These comprehensive data argue that intrinsic factors shape niche breadth variation in microbes.

show abstract

Section: Human Yeast Pathogens Include Both Carbon Generalists and Sp...mentioning

confidence: 99%

Genomic factors shape carbon and nitrogen metabolic niche breadth across Saccharomycotina yeasts

Opulente,

LaBella,

Harrison

et al. 2024

Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…Supervised models rely on labeled training data as input [e.g., random forest (RF), logistic regression (LR)], whereas unsupervised ones use unlabeled raw data (e.g., neural network, hidden Markov model) to enable predictions/classifications given sufficient data. Machine learning models have been successfully used to predict both genomic features ( 31 ) and phenotypic traits ( 32 – 34 ) and can be used with any type of biological, including genomic, data ( 35 ).…”

Section: Introductionmentioning

confidence: 99%

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Riedling,

Walker,

Rokas

2024

Microbiol Spectr

Self Cite

View full text Add to dashboard Cite

Fungal secondary metabolites (SMs) contribute to the diversity of fungal ecological communities, niches, and lifestyles. Many fungal SMs have one or more medically and industrially important activities (e.g., antifungal, antibacterial, and antitumor). The genes necessary for fungal SM biosynthesis are typically located right next to each other in the genome and are known as biosynthetic gene clusters (BGCs). However, whether fungal SM bioactivity can be predicted from specific attributes of genes in BGCs remains an open question. We adapted machine learning models that predicted SM bioactivity from bacterial BGC data with accuracies as high as 80% to fungal BGC data. We trained our models to predict the antibacterial, antifungal, and cytotoxic/antitumor bioactivity of fungal SMs on two data sets: (i) fungal BGCs (data set comprised of 314 BGCs) and (ii) fungal (314 BGCs) and bacterial BGCs (1,003 BGCs). We found that models trained on fungal BGCs had balanced accuracies between 51% and 68%, whereas training on bacterial and fungal BGCs had balanced accuracies between 56% and 68%. The low prediction accuracy of fungal SM bioactivities likely stems from the small size of the data set; this lack of data, coupled with our finding that including bacterial BGC data in the training data did not substantially change accuracies currently limits the application of machine learning approaches to fungal SM studies. With >15,000 characterized fungal SMs, millions of putative BGCs in fungal genomes, and increased demand for novel drugs, efforts that systematically link fungal SM bioactivity to BGCs are urgently needed. IMPORTANCE Fungi are key sources of natural products and iconic drugs, including penicillin and statins. DNA sequencing has revealed that there are likely millions of biosynthetic pathways in fungal genomes, but the chemical structures and bioactivities of >99% of natural products produced by these pathways remain unknown. We used artificial intelligence to predict the bioactivities of diverse fungal biosynthetic pathways. We found that the accuracies of our predictions were generally low, between 51% and 68%, likely because the natural products and bioactivities of only very few fungal pathways are known. With >15,000 characterized fungal natural products, millions of putative biosynthetic pathways present in fungal genomes, and increased demand for novel drugs, our study suggests that there is an urgent need for efforts that systematically identify fungal biosynthetic pathways, their natural products, and their bioactivities.

show abstract

Model-Driven Elucidation of Lactose and Galactose Metabolism via Oxidoreductive Pathway inSungouiella intermediafor Cell Factory Applications

Peri,

Domenzain,

Alalam

et al. 2024

Preprint

View full text Add to dashboard Cite

Converting industrial side streams into value-added chemicals using microbial cell factories is of increasing interest, as such processes offer solutions to reduce waste and production costs. However, developing new, efficient cell factories for precision fermentation remains challenging due to limited knowledge about their metabolic capabilities. Here, we investigate the lactose and galactose metabolism of the non-conventional yeast Sungouiella intermedia (formerly Candida intermedia), using knowledge-matching of high-quality genome-scale metabolic model (GEM) with extensive experimental analysis and determine its potential as a future cell factory on lactose-rich industrial side-streams. We show that this yeast possesses the conserved Leloir pathway as well as an oxidoreductive galactose catabolic route. Contextualization of RNAseq data into Sint-GEM highlights the regulatory mechanisms on the oxidoreductive pathway and how this pathway can enable adaptation to diverse environments. Model simulations, together with experimental data from continuous and batch bioreactors, indicate that S. intermedia uses upstream enzymes of the oxidoreductive pathway, in a condition-dependent manner, and produce the sugar alcohol galactitol as a carbon overflow metabolite, coupled to redox co-factor balancing during both lactose and galactose growth. Furthermore, the new metabolic insights facilitated the development of an improved bioprocess design, where an engineered S. intermedia strain could achieve galactitol yields of >90% of the theoretical maximum at improved production rates using the industrial side-stream cheese whey permeate as feedstock. Additional strain engineering resulted in galactitol-to-tagatose conversion, proving the versatility of the future production host. Overall, this work sheds new light on the intrinsic interplay between parallel metabolic pathways that shape the lactose and galactose catabolism in S. intermedia. It also demonstrates how a GEM combined with experimental analysis can work in synergy to fast-forward metabolic characterization and development of new, non-conventional yeast cell factories.

show abstract

Machine learning illuminates how diet influences the evolution of yeast galactose metabolism

Cited by 3 publications

References 58 publications

Genomic factors shape carbon and nitrogen metabolic niche breadth across Saccharomycotina yeasts

Genomic factors shape carbon and nitrogen metabolic niche breadth across Saccharomycotina yeasts

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Model-Driven Elucidation of Lactose and Galactose Metabolism via Oxidoreductive Pathway inSungouiella intermediafor Cell Factory Applications

Contact Info

Product

Resources

About