Machine learning for optimal growth temperature prediction of prokaryotes using amino acid descriptors
Colette, S.; Francois, J.; De Moor, B.; van Noort, V.
Show abstract
MotivationThe optimal growth temperature (OGT) of organisms is valuable in bioprospecting enzymes that work under extreme conditions. Existing OGT prediction models achieve high accuracy, but mainly capture trends of overrepresented groups in the training set including organisms that thrive at moderate temperatures and those from well-described taxa. ResultsIn this study, we incorporated weighted scoring and phylogenetic splits to improve the generalizability of the prediction models. We first built a new growth temperature dataset comprising more than 21,000 species distributed over all three domains of life, with special attention to include OGT and extreme temperature data. We then trained machine learning models on the OGT data of 6,401 prokaryotes using proteome-averaged amino acid descriptors. The best-performing model was the multilayer perceptron with a cross-validated RMSE of 5.07{degrees}C ({+/-} 0.24) and an R2 of 0.89 ({+/-} 0.04). The most important proteome features were related to backbone flexibility, charged residues, as well as surface accessibility. Availability and ImplementationThe MLP model is integrated in the command line tool OGTFinder and available under MIT license at: https://github.com/SC-Git1/OGTFinder.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- A large-scale genome-based survey of acidophilic Bacteria suggests that genome streamlining is an adaption for life at low pH 95%
- FeGenie: a comprehensive tool for the identification of iron genes and iron gene neighborhoods in genomes and metagenome assemblies 93%
- Machine learning uncovers a data-driven transcriptional regulatory network for the Crenarchaeal thermoacidophile Sulfolobus acidocaldarius 93%
Similar papers in this journal
- Robust genome-based delineation of bacterial genera 95%
- Evolutionary history of dimethylsulfoniopropionate (DMSP) demethylation enzyme DmdA in marine bacteria 93%
- Non-synonymous to synonymous substitutions suggest that orthologs tend to keep their functions, while paralogs are a source of functional novelty 93%
Similar papers in this journal
- EcoFoldDB: Protein structure-guided functional profiling of ecologically relevant microbial traits at the metagenome scale 92%
- Cellular life from the three domains and viruses are transcriptionally active in a hypersaline desert community 92%
- Metabolic potential and survival strategies of microbial communities across extreme temperature gradients on Deception Island volcano, Antarctica 92%
Similar papers in this journal
- Genomic and evolutionary factors influencing the prediction accuracy of optimal growth temperature in prokaryotes 94%
- Selection for translational efficiency in genes associated with alphaproteobacterial gene transfer agents 93%
- Decomposing a San Francisco Estuary microbiome using long read metagenomics reveals species and species- and strain-level dominance from picoeukaryotes to viruses 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.