Skip to main content

Overview

TRIFID provides pre-computed predictions and predictive features for a wide range of species and genome assemblies. Each dataset includes functionality scores for all annotated splice isoforms in the respective genome version.

Pre-computed Predictions

The following table lists all available TRIFID predictions organized by genome assembly, species, and annotation source:
Genome AssemblySpeciesCommon NameModelVersionDatabaseRelease DateData Access
GRCh38Homo sapiensHumanHumanv1GENCODE27 - 08.2017Download
GRCh38Homo sapiensHumanHumanv2GENCODE42 - 04.2022Download
GRCh38Homo sapiensHumanHumanv2GENCODE37 - 02.2021Download
GRCh38Homo sapiensHumanRefSeqv2RefSeq - NCBI110 - 02.2020Download
GRCh37Homo sapiensHumanRefSeqv2RefSeq - NCBI105 - 02.2020Download
GRCh37Homo sapiensHumanHumanv2GENCODE19 - 12.2013Download
GRCm39Mus musculusMouseMousev2GENCODE31 - 04.2022Download
GRCm38Mus musculusMouseMousev2GENCODE25 - 11.2019Download
mRatBN7.2Rattus norvegicusRatVertebratesv2Ensembl105 - 12.2021Download
GRCz11Danio rerioZebrafishVertebratesv2Ensembl104 - 05.2021Download
GRCg7bGallus gallusChickenVertebratesv2Ensembl108 - 10.2022Download
Pan_tro_3.0Pan troglodytesChimpanzeeVertebratesv2Ensembl104 - 05.2021Download
Sscrofa11.1Sus scrofaPigVertebratesv2Ensembl108 - 10.2022Download
ARS-UCD1.2Bos taurusCowVertebratesv2Ensembl104 - 05.2021Download
Mmul_10Macaca mulattaMacaqueVertebratesv2Ensembl105 - 12.2021Download
BDGP6Drosophila melanogasterFruitflyInvertebratesv2Ensembl - Flybase107 - 07.2022Download
WBcel235Caenorhabditis elegansWormInvertebratesv2Ensembl - Wormbase108 - 10.2022Download

Model Versions

v1 (Initial Release)

The initial TRIFID model trained on GENCODE Release 27 (GRCh38.p10) with 45 predictive features. Released in March 2021.

v2 (Current Release)

The updated TRIFID model with 47 predictive features (2 additional features added). Released in September 2022 with support for multiple species and genome assemblies.

Dataset Contents

Each sharepoint contains:
  • Predictions file: TRIFID functionality scores for all isoforms
  • Features file: All 47 predictive features used by the model
  • Metadata: Gene and transcript annotations

File Format

Prediction files are provided as compressed TSV files (trifid_predictions.tsv.gz) with the following columns:
  • transcript_id: Ensembl transcript identifier
  • gene_id: Ensembl gene identifier
  • gene_name: Gene symbol
  • trifid_score: Functionality probability (0-1)
  • appris: APPRIS annotation label
  • Additional feature columns

Requesting New Data

If you need predictions for a specific genome version or species not listed above, please open an issue on the GitHub repository.

Build docs developers (and LLMs) love