OGTFinder: Optimal Growth Temperatures for Over 15,000 Prokaryotic Species

A curated data set of optimal growth temperatures (OGT) for more than 15,000 species across all three domains of life, with special attention to extreme temperatures. The data was compiled by Sophie Colette and last updated in April 2026. It was used to train a machine learning model for OGT prediction, achieving a test RMSE of 5.49 °C.

Use Cases

• Predicting optimal growth temperatures for bacteria and archaea based on proteome-averaged amino acid descriptors.
• Training machine learning models for enzyme bioprospecting using features related to backbone flexibility and charged residues.
• Improving model generalizability for underrepresented taxa using weighted scoring and phylogenetic splits.
• Analyzing relationships between proteome features, surface accessibility, and organism growth temperature.

Strengths

• Contains data for more than 15,000 species distributed over all three domains of life.
• Includes special attention to OGT and extreme temperature data.
• Associated ML model achieved a test RMSE of 5.49 °C and an R² of 0.84.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Sophie Colette via figshare.

Collection Method

Curated data set compiled for machine learning model training.

Time Range

null

Freshness

Last updated 2026-04-13 07:36:28; freshness should be verified.

Geography

null