{
    "io_domain": {
        "input_subdomain": [
            {
                "uri": {
                    "uri": "https://data.glygen.org/ln2data/releases/data/v-1.4.5/source/uniprot-proteome-10116-rattus-norvegicus-2019-05.nt",
                    "filename": "uniprot-proteome-10116-rattus-norvegicus-2019-05.nt"
                }
            },
            {
                "uri": {
                    "uri": "https://data.glygen.org/ln2data/releases/data/v-1.4.5/reviewed/rat_protein_idmapping.csv",
                    "filename": "rat_protein_idmapping.csv"
                }
            }
        ],
        "output_subdomain": [
            {
                "uri": {
                    "uri": "http://data.glygen.org/ln2data/releases/data/v-1.4.5/reviewed/rat_protein_xref_interpro.csv",
                    "filename": "rat_protein_xref_interpro.csv"
                },
                "mediatype": "csv"
            },
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "http://data.glygen.org/ln2data/releases/data/v-1.4.5/reviewed/rat_protein_xref_interpro.stat.csv",
                    "filename": "rat_protein_xref_interpro.stat.csv"
                }
            }
        ]
    },
    "error_domain": {},
    "spec_version": "1.3.0",
    "object_id": "https://biocomputeobject.org/GLY_000243/v-1.4.5",
    "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": "Rattus Norvegicus",
                        "category_name": "species"
                    },
                    {
                        "category_value": "Protein",
                        "category_name": "molecule"
                    },
                    {
                        "category_value": "Interpro cross-references",
                        "category_name": "tags"
                    },
                    {
                        "category_value": "csv",
                        "category_name": "file_type"
                    },
                    {
                        "category_value": "Reviewed",
                        "category_name": "status"
                    },
                    {
                        "category_value": "internal",
                        "category_name": "scope"
                    }
                ]
            }
        }
    ],
    "provenance_domain": {
        "license": "https://creativecommons.org/licenses/by/4.0/",
        "version": "v-1.4.5",
        "name": "Rat 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": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Xavier Holmes",
                "email": "xavierh@email.gwu.edu"
            },
            {
                "affiliation": "EMBL-EBI-UniProt",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Maria Martin",
                "email": "martin@ebi.ac.uk"
            },
            {
                "affiliation": "EMBL-EBI-UniProt",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Preethi Vasudev",
                "email": "pvasudev@ebi.ac.uk"
            }
        ],
        "created": "2019-09-06T14:06:38.970158"
    },
    "description_domain": {
        "keywords": [
            "protein family",
            "protein domain",
            "protein functional site",
            "InterPro",
            "cross-reference"
        ],
        "pipeline_steps": [
            {
                "name": "make-protein-dataset.py",
                "step_number": 1,
                "version": "",
                "input_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.4.5/reviewed/rat_protein_idmapping.csv",
                        "filename": "rat_protein_idmapping.csv"
                    },
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.4.5/source/uniprot-proteome-10116-rattus-norvegicus-2019-05.nt",
                        "filename": "uniprot-proteome-10116-rattus-norvegicus-2019-05.nt"
                    }
                ],
                "output_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.4.5/reviewed/rat_protein_xref_interpro.csv",
                        "filename": "rat_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 and extracts the data as a csv dataset"
            }
        ]
    },
    "execution_domain": {
        "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/GW-HIVE/glygen-backend-integration/blob/master/pipeline/integrator/make-protein-dataset.py",
                    "filename": "make-protein-dataset.py"
                }
            }
        ]
    },
    "usability_domain": [
        "The Rat UniprotKB Xref InterPro contains rat [taxid:10116] 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 2019-05 UniProt release. If you use this dataset please provide proper attribution to EMBL-EBI-UniProtKB and GlyGen"
    ],
    "etag": "",
    "_id": {
        "$oid": "5d72a02ef5c47b0f43423364"
    }
}
