{
    "object_id": "https://biocomputeobject.org/GLY_000961/v-2.8.1",
    "spec_version": "1.3.0",
    "etag": "3bc93b9a0dce8a73c0131933d4a07fc9",
    "provenance_domain": {
        "name": "Human Glycosylation Sites ccRCC (PDC)",
        "version": "v-2.8.1",
        "license": "https://creativecommons.org/licenses/by/4.0/",
        "created": "2024-07-23T00:39:27",
        "modified": "2024-07-23T00:40:06.813Z",
        "contributors": [
            {
                "name": "Jeet Vora",
                "affiliation": "The George Washington University",
                "email": "jeetvora@gwu.edu",
                "contribution": [
                    "createdBy"
                ]
            },
            {
                "name": "Karina Martinez",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "karinamartinez@gwu.edu"
            },
            {
                "name": "Robel Kahsay",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "rykahsay@gwu.edu"
            },
            {
                "name": "Jingyue Wu",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "jingyue.wu@email.gwu.edu"
            },
            {
                "name": "Luke Johnson",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "luke.johnson@email.gwu.edu"
            },
            {
                "name": "Urnisha Bhuiyan",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "sbhuiyan28@email.gwu.edu"
            },
            {
                "name": "Cyrus Au-Yeung",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "cyruschauyeung@gwu.edu"
            },
            {
                "name": "Kate Warner",
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "email": "k.warner1@gwu.edu"
            }
        ]
    },
    "usability_domain": [
        "The Human Glycosylation Sites (PDC) dataset contains glycosylation information and intact glycopeptide abundances from Homo sapiens (taxid:9606) clear cell renal cell carcinoma (ccRCC) tumor and normal adjacent tissues collected by the CPTAC program. The processed data was retrieved from the Proteomic Data Commons (PDC) and can be accessed at https://proteomic.datacommons.cancer.gov/pdc/study/PDC000471. GlyTouCan and UniProtKB accessions were mapped by the GlyGen team to the glycan compositions and gene names provided by the authors, respectively. If you use this dataset please provide proper attribution to the source publication[1], PDC and GlyGen.",
        "References: 1) Lih TM, Cho KC, Schnaubelt M, Hu Y, Zhang H. Integrated glycoproteomic characterization of clear cell renal cell carcinoma. Cell Rep. 2023 May 30;42(5):112409. doi: 10.1016/j.celrep.2023.112409. Epub 2023 Apr 18. PMID: 37074911; PMCID: PMC10247542."
    ],
    "description_domain": {
        "pipeline_steps": [
            {
                "name": "make-proteoform-dataset.py",
                "step_number": 1,
                "input_list": [],
                "description": "The script performs QC steps and creates a proteoform dataset according to the GlyGen data model.",
                "prerequisite": [],
                "output_list": []
            }
        ],
        "keywords": [
            "N-linked glycosylation",
            "glycoproteomics",
            "clear cell renal cell carcinoma",
            "PDC"
        ]
    },
    "parametric_domain": [
        {
            "value": ""
        }
    ],
    "io_domain": {
        "output_subdomain": [
            {
                "uri": {
                    "uri": "https:/data.glygen.org/ln2releases/data/v-2.8.1/reviewed/human_proteoform_glycosylation_sites_pdc_ccrcc.csv",
                    "filename": "human_proteoform_glycosylation_sites_pdc_ccrcc.csv"
                },
                "mediatype": "csv"
            },
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "https:/data.glygen.org/ln2releases/data/v-2.8.1/reviewed/human_proteoform_glycosylation_sites_pdc_ccrcc.stat.csv",
                    "filename": "human_proteoform_glycosylation_sites_pdc_ccrcc.stat.csv"
                }
            }
        ],
        "input_subdomain": [
            {
                "uri": {
                    "filename": "human_protein_masterlist.csv",
                    "uri": "https:/data.glygen.org/ln2releases/data/v-2.8.1/reviewed/human_protein_masterlist.csv"
                }
            },
            {
                "uri": {
                    "filename": "n_sequon_info.csv",
                    "uri": "https:/data.glygen.org/ln2releases/data/v-2.8.1/misc/n_sequon_info.csv"
                }
            },
            {
                "uri": {
                    "filename": "pdc_glytoucan_mapping.csv",
                    "uri": "https:/data.glygen.org/ln2releases/data/v-2.8.1/misc/pdc_glytoucan_mapping.csv"
                }
            },
            {
                "uri": {
                    "filename": "ccRCC_TMT_intact_glycopeptide_abundance_MD-MAD.tsv",
                    "uri": "https:/data.glygen.org/ln2downloads/pdc/2024_04_18/ccRCC_TMT_intact_glycopeptide_abundance_MD-MAD.tsv"
                }
            }
        ]
    },
    "execution_domain": {
        "script": [
            {
                "uri": {
                    "uri": "https://github.com/glygener/glygen-backend-integration/blob/master/dataset-maker/make-proteoform-dataset.py",
                    "filename": "make-proteoform-dataset.py"
                }
            }
        ],
        "script_driver": "manual",
        "software_prerequisites": [
            {
                "version": "2.7.5",
                "name": "Python",
                "uri": {
                    "uri": "https://www.python.org/download/releases/2.7.5/"
                }
            }
        ],
        "external_data_endpoints": [],
        "environment_variables": {}
    },
    "extension_domain": [
        {
            "extension_schema": "https://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": "Homo sapiens",
                        "category_name": "species"
                    },
                    {
                        "category_value": "Proteoform",
                        "category_name": "molecule"
                    },
                    {
                        "category_value": "Glycosylation Sites ccRCC (PDC)",
                        "category_name": "tag"
                    },
                    {
                        "category_value": "csv",
                        "category_name": "file_type"
                    },
                    {
                        "category_value": "Reviewed",
                        "category_name": "status"
                    },
                    {
                        "category_value": "internal",
                        "category_name": "scope"
                    }
                ]
            }
        }
    ],
    "error_domain": {}
}