The Template Predictor API runs BLAST, SparksX, and HHsearch. It will output a HHR sequence alignment file, one alignment file from hhsearch and one from sparksx, and a PSSM. It will also output a TSV which is a parsed version of these files.

1 Quickstart
- 1.1 Command Line Examples
- 1.2 Python Examples
2 Inputs
3 Outputs
4 Notes
- 4.1 Output File interpretation

Quickstart

Command Line Examples

Submit a template predictor job:

cyrus engine submit template-predictor SECVENGGFCPDPEKMGDWCCGRCIRNECRNG

Python Examples

Submit a template predictor job:

from engine.template-predictor.client import TemplatePredictorClient

client = TemplatePredictorClient()
job_id = client.submit(sequence="SECVENGGFCPDPEKMGDWCCGRCIRNECRNG")

Inputs

--sequence (str)
- Sequence to run template predictor on

Outputs

sparksx-cambyses-aln.json
- sparksX alignment file
hhsearch-cambyses-aln.json
- HHsearch alignment file
sequence.hhr
- .hhr alignment file
target.pssm
- BLAST PSSM file
results.tsv
- TSV format of results parsed from hhsearch and sparksX

Notes

Output File interpretation

results.tsv
- homology_probability
  - <60%: low confidence in homology modeling
  - 60-80%: major modeling required
  - >80%: high confidence in homology modeling
- template_coverage
  - <50%: homology modeling will lead to divergent models
  - 50-70%: homology modeling will lead to fairly converged models
  - >70%: homology modeling will lead to very converged models
- sequence_identity
- gaps_over_12_residues
  - The number of gaps greater than 12 residues long in each template
- positions_in_no_templates
  - A list of all positions which are missing from each template