GTalign-web API help

Table of contents

1. Submitting query structure(s)
2. Specifying search databases
3. Specifying search settings
4. Checking job status
5. Downloading the results

GTalign-web allows submitting queries and checking their status using command-line interface.

Submitting query structure(s)

One or more structure files can be submitted using the following command:

curl https://bioinformatics.lt/comer/gtalign/api/submit -F input_query_files=@first_structure.pdb -F input_query_files=@second_structure.cif

To submit for a multimer search using GTcomplex, use a different URL:

curl https://bioinformatics.lt/comer/gtalign/api/submit_complex -F input_query_files=@multimeric_structure_file.cif

The server responds in JSON format:

The field "success" indicates if there are any errors, and "job_id" is the identifier of the GTaling-web job. This identifier can be used in further requests or for opening the job in a browser. "MyJobID" will be used as job identifier in the following text.

Specifying search databases

Search is possible using the same databases that are available in the web interface. The list of available databases for monomeric protein search can be cheched using the following command:

curl https://bioinformatics.lt/comer/gtalign/api/available_databases_protein

The response contains database descriptions and names for querying the server:

For example, the following command submits a query to search in the UniProtKB/SwissProt database:

curl https://bioinformatics.lt/comer/gtalign/api/submit -F input_query_files=@structure.pdb -F database=swissprot_v4

Similarly, available databases can be queried for a GTcomplex search:

curl https://bioinformatics.lt/comer/gtalign/api/available_databases_complex

Specifying search settings

Search settings can be specified using the following keywords:

• s: TM-score threshold [0, 1)
• sort: sorting options (integer from 0 to 8)
• nhits: number of reported hits (max 1000)
• presimilarity: threshold for sequence similarity pre-screening [0, 100]
• prescore: threshold for structure similarity pre-screening [0, 1)
• speed: search speed optimization (integer from 0 to 13, the larger, the faster)
• nogaps: possibility to exclude deletion positions (gaps in the query) from alignments (true, false)

The following command runs a search using slower speed and pre-filtering of queries having sequence similarity of at least 20%, reporting only hits having TM-score > 0.6:

curl https://bioinformatics.lt/comer/gtalign/api/submit -F input_query_files=@structure.pdb -F s=0.6 -F presimilarity=20 -F speed=9

More details and the default parameters can be found in the main Help page.

Checking job status

The job status can be checked after submitting it:

curl https://bioinformatics.lt/comer/gtalign/api/job_status/MyJobID

The status of the job is reported in JSON format. First, the job is "new", the "queued", then "running":

For finished jobs, number of successful results is reported together with links to the job's web interface and for downloading of outputs:

Downloading the results

Downloading results files

All job data (input and output) can be downloaded using the download URL:

curl https://bioinformatics.lt/comer/gtalign/download/results/MyJobID

The same URL can be used to download JSON outputs for particular queries by specifying the 0-based index of the query at the end:

curl https://bioinformatics.lt/comer/gtalign/download/results/MyJobID/0

Dowloading the input structures

All input structures in a TAR archive can be downloaded as follows:

curl https://bioinformatics.lt/comer/gtalign/download/input/MyJobID

GTalign splits multi-chain and multi-model structures to separate querys. Similarly to results for a single query, it's input structure in PDB format can be also downloaded:

curl https://bioinformatics.lt/comer/gtalign/download/input/MyJobID/0