{
    "io_domain": {
        "input_subdomain": [
            {
                "uri": {
                    "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/source/uniprot-proteome-10090-mus-musculus-2018-11.nt",
                    "filename": "uniprot-proteome-10090-mus-musculus-2018-11.nt"
                }
            }
        ],
        "output_subdomain": [
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/reviewed/mouse_protein_xref_interpro.csv",
                    "filename": "mouse_protein_xref_interpro.csv"
                }
            },
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/reviewed/mouse_protein_xref_interpro.stat.csv",
                    "filename": "mouse_protein_xref_interpro.stat.csv"
                }
            }
        ]
    },
    "spec_version": "1.3.0",
    "object_id": "https://biocomputeobject.org/GLY_000107/v-1.0.13",
    "extension_domain": [
        {
            "extension_schema": "http://www.w3id.org/biocompute/extension_domain/1.1.0/dataset/dataset_extension.json",
            "dataset_extension": {
                "additional_license": {
                    "data_license": "https://creativecommons.org/licenses/by/4.0/",
                    "script_license": "https://www.gnu.org/licenses/gpl-3.0.en.html"
                },
                "dataset_categories": [
                    {
                        "category_value": "Reviewed",
                        "category_name": "status"
                    }
                ]
            }
        }
    ],
    "provenance_domain": {
        "license": "https://creativecommons.org/licenses/by/4.0/",
        "name": "Mouse UniProtKB Xref InterPro",
        "contributors": [
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Jeet Vora",
                "email": "jeetvora@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Rahi Navelkar",
                "email": "rsn13@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Robel Kahsay",
                "email": "rykahsay@gwu.edu"
            },
            {
                "affiliation": "EMBL-EBI-UniProt",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Maria Martin",
                "email": "martin@ebi.ac.uk"
            },
            {
                "affiliation": "EMBL-EBI-UniProt",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Leyla Garcia",
                "email": "ljgarcia@ebi.ac.uk"
            },
            {
                "affiliation": "EMBL-EBI-UniProt",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Preethi Vasudev",
                "email": "pvasudev@ebi.ac.uk"
            }
        ],
        "created": "2019-03-15T12:05:44.324142",
        "modified": "2019-04-23T16:59:28.330806",
        "version": "v-1.0.13"
    },
    "description_domain": {
        "keywords": [
            "protein family",
            "protein domain",
            "protein functional site"
        ],
        "pipeline_steps": [
            {
                "name": "extract-from-endpoint.py",
                "step_number": 1,
                "version": "1.0",
                "input_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/source/uniprot-proteome-10090-mus-musculus-2018-11.nt",
                        "filename": "uniprot-proteome-10090-mus-musculus-2018-11.nt"
                    },
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/reviewed/mouse_protein_idmapping.csv",
                        "filename": "mouse_protein_idmapping.csv"
                    }
                ],
                "output_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.0.13/reviewed/mouse_protein_xref_interpro.csv",
                        "filename": "mouse_protein_xref_interpro.csv"
                    }
                ],
                "prerequisite": [],
                "description": "The script loads the UniProtKB accessions that are cross-referenced to InterPro database accessions/identifiers from the triplestore containing triples from nt file and maps to the UniProtKB canonical accessions extracts the data as a csv dataset"
            }
        ]
    },
    "execution_domain": {
        "external_data_endpoints": [],
        "environment_variables": {},
        "script_driver": "manual",
        "software_prerequisites": [
            {
                "version": "2.7.5",
                "name": "Python",
                "uri": {
                    "uri": "https://www.python.org/download/releases/2.7.5/"
                }
            }
        ],
        "script": [
            {
                "uri": {
                    "uri": "https://github.com/glygener/glygen-backend-integration/blob/master/pipeline/extractor/extract-from-endpoint.py",
                    "filename": "extract-from-endpoint.py"
                }
            }
        ]
    },
    "error_domain": {
        "empirical_error": {},
        "algorithmic_error": {},
        "emperical_error": {}
    },
    "parametric_domain": [
        {
            "step": "",
            "value": "",
            "param": ""
        }
    ],
    "usability_domain": [
        "The Mouse UniprotKB Xref InterPro contains mouse [taxid:10090] UniProtKB canonical accessions cross-referenced to InterPro database accessions/identifiers",
        "InterPro is a database of protein families, domains and functional sites in which identifiable features found in known proteins can be applied to new protein sequences in order to functionally characterise them",
        "The dataset is derived from 2018-11 UniProt release. If you use this dataset please provide proper attribution to EMBL-EBI-UniProt and GlyGen"
    ],
    "etag": "6d53b987273d19d058c5ccb6a1fd7dc8",
    "_id": {
        "$oid": "5c8bc84bf5c47b4a20d22e6b"
    }
}
