Download Data

Access the complete tAI datasets in CSV format for your research. All data is freely available for academic use.

📄 Eukaryota tAI (NCBI RefSeq)

tRNA gene copy numbers and dynamic gtAI optimality profiles derived from Eukaryotic genome assemblies, processed through tRNAscan-SE and capped with our Interquartile Range (IQR) algorithm to rescue mammalian SINE contamination.

📊 1,292 species 📐 61 codons 📁 CSV
Download eukaryota_tai.csv

📄 Bacteria tAI (NCBI RefSeq)

Bacterial optimality profiles processed with domain-specific wobble constraints, including strictly limiting Inosine editing to the TadA context (tRNA-Arg) and modeling U34 hypermodifications.

📊 6,996 species 📐 61 codons 📁 CSV
Download bacteria_tai.csv

📄 Archaea tAI (NCBI RefSeq)

Archaeal optimality profiles incorporating unique clade-specific modifications such as Agmatidine editing at the Isoleucine wobble position.

📊 434 species 📐 61 codons 📁 CSV
Download archaea_tai.csv

📄 Full Tree of Life (Merged)

The complete combined dataset of all Eukaryotic, Bacterial, and Archaeal species in a single flat file. Includes raw Wi metrics and dynamically derived gtAI scaling values.

📊 8,722 species 📐 61 codons 📁 JSON
Download species_tai_gtai.json

📄 Vertebrates (NCBI RefSeq)

Highly-curated subset of the vertebrate clade. SINE contamination explicitly capped using the IQR algorithm.

📊 578 species 📐 61 codons 📁 CSV
Download vertebrate_tai.csv

📄 Mammals (NCBI RefSeq)

All mammalian genomes in the database, with corrected B1/B2 retrotransposon tRNA inflation.

📊 157 species 📐 61 codons 📁 CSV
Download mammals_tai.csv

📄 Birds (NCBI RefSeq)

Avian genomes across the RefSeq database.

📊 106 species 📐 61 codons 📁 CSV
Download birds_tai.csv

📄 Fish (NCBI RefSeq)

Actinopterygii and Chondrichthyes genomes across the database.

📊 232 species 📐 61 codons 📁 CSV
Download fish_tai.csv

Data Format

All files follow the same CSV structure

Column Structure

Column Description Example
Species Binomial species name (genus + epithet) Homo sapiens
AAA Wi weight for the AAA (Lys) codon 12.0
AAC Wi weight for the AAC (Asn) codon 25.0
... 61 codon columns total (all 61 sense codons, excluding stop codons UAA, UAG, UGA) ...
UUU Wi weight for the UUU (Phe) codon 10.0

Important Notes

  • Values represent raw tRNA gene copy numbers (Wi) incorporating wobble base-pairing optimization, not normalized tAI scores
  • To compute tAI for a gene, normalize each Wi by dividing by max(Wi), then take the geometric mean across all codons in the coding sequence
  • Codons use RNA notation (U instead of T)
  • The three stop codons (UAA, UAG, UGA) are excluded
  • AUG (Met, start codon) is included as it also serves as an internal codon
  • Some values may be fractional due to wobble-pairing weight redistribution

Citation

If you use tAIatlas data in your research, please cite:

tAIatlas: A Universal tRNA Adaptation Index Database

tAIatlas Project (2025). tAIatlas: A Universal tRNA Adaptation Index Database. Available at: https://taiatlas.github.io

Original tAI methodology

dos Reis, M., Savva, R., & Wernisch, L. (2004). Solving the riddle of codon usage preferences: a test for translational selection. Nucleic Acids Research, 32(17), 5036–5044. doi:10.1093/nar/gkh834