Dtu Protein Tools

Configuration File: dtu_protein_tools.json Tool Type: Local Tools Count: 1

This page contains all tools defined in the dtu_protein_tools.json configuration file.

Available Tools

dtu_protein_predict (Type: DTUProteinTool)

Run DTU Health Tech machine-learning protein predictors on a protein sequence via the BioLib clou…

dtu_protein_predict tool specification

Tool Information:

  • Name: dtu_protein_predict

  • Type: DTUProteinTool

  • Description: Run DTU Health Tech machine-learning protein predictors on a protein sequence via the BioLib cloud runner (pybiolib). Choose a model: ‘deeptmhmm’ (DTU/DeepTMHMM transmembrane topology -> per-residue topology string of inside/outside/TM-helix/beta-strand/signal plus TM region table), ‘signalp’ (DTU/SignalP_6 signal-peptide detection + cleavage site), or ‘deeploc’ (KU/DeepLocPro subcellular localization, prokaryotic). Input is a protein FASTA: pass ‘sequence’/’fasta’ as inline FASTA text or a bare amino-acid string, or ‘fasta_path’ as a path to a .fasta file. IMPORTANT: jobs run on the shared BioLib cloud (anonymously, no account needed) and commonly take 2-5 minutes; raise ‘max_wait_time’ for many/large sequences. NOTE: the original eukaryotic DeepLoc 2.0 is not published on BioLib; ‘deeploc’ maps to the prokaryotic DeepLocPro model, whose upstream BioLib app currently crashes during embedding (device-mismatch) and returns a clean error until fixed – ‘deeptmhmm’ and ‘signalp’ are the reliably working models. Prerequisite: pip install pybiolib. Use for: predicting membrane protein topology, finding signal peptides/secretion signals, and bacterial protein subcellular location.

Parameters:

  • model (string) (optional) Which predictor to run: ‘deeptmhmm’ (transmembrane topology), ‘signalp’ (signal peptide), or ‘deeploc’ (prokaryotic subcellular localization).

  • sequence (string) (optional) Protein input as inline FASTA text (one or more ‘>header’ records) or a bare single-letter amino-acid string (it will be wrapped as ‘>query’). Mutually exclusive with fasta_path. Example: ‘MELAALCRWGLLLALLPPGAAS’.

  • fasta (string) (optional) Alias for ‘sequence’. Inline FASTA text or a bare amino-acid string.

  • fasta_path (string) (optional) Path to a local FASTA (.fasta/.fa) file containing one or more protein records. Use instead of ‘sequence’ when the input is already on disk.

  • max_wait_time (integer) (optional) Maximum seconds to wait for the BioLib cloud job to finish (default 600, min 30, max 1800). Increase for many/long sequences.

Example Usage:

query = {
    "name": "dtu_protein_predict",
    "arguments": {
    }
}
result = tu.run(query)