{
    "bcoid": "GLY_000573",
    "title": "Mouse Proteoform Citations (Literature Mining)",
    "description": "The Mouse Proteoform Citations [Automatic Literature Mining] dataset contains Mus musculus [taxid:10090] ...",
    "filename_list": "mouse_proteoform_citations_glycosylation_sites_literature_mining.csv",
    "filename": "mouse_proteoform_citations_glycosylation_sites_literature_mining.csv",
    "filetype": "csv",
    "categories": {
        "species": [
            "Mus musculus"
        ],
        "molecule": [
            "Proteoform"
        ],
        "tag": [
            "Protein Citations"
        ],
        "file_type": [
            "csv"
        ],
        "status": [
            "Reviewed"
        ],
        "scope": [
            "internal"
        ],
        "file_category": [
            "citations"
        ]
    },
    "minitable": {
        "content": [
            [
                "O55188-1...",
                "Glycosylatio..."
            ],
            [
                "O55188-1...",
                "Glycosylatio..."
            ]
        ],
        "headers": [
            "uniprotkb_ca...",
            "title..."
        ],
        "colwidth": [
            "40%",
            "20%"
        ]
    },
    "sampledata": {
        "type": "table",
        "data": [
            [
                {
                    "type": "string",
                    "label": "uniprotkb_canonical_ac"
                },
                {
                    "type": "string",
                    "label": "title"
                },
                {
                    "type": "string",
                    "label": "journal_name"
                },
                {
                    "type": "string",
                    "label": "publication_date"
                },
                {
                    "type": "string",
                    "label": "authors"
                },
                {
                    "type": "string",
                    "label": "xref_key"
                },
                {
                    "type": "string",
                    "label": "xref_id"
                },
                {
                    "type": "string",
                    "label": "src_xref_key"
                },
                {
                    "type": "string",
                    "label": "src_xref_id"
                },
                {
                    "type": "string",
                    "label": "glytoucan_ac"
                },
                {
                    "type": "string",
                    "label": "isoform_mapping_notes"
                }
            ],
            [
                "O55188-1",
                "Glycosylation of Dentin Matrix Protein 1 is critical for osteogenesis.",
                "Scientific reports",
                "2015 Dec 4",
                "Sun Y, Weng Y, Zhang C, Liu Y, Kang C, Liu Z, Jing B, Zhang Q, Wang Z",
                "protein_xref_pubmed",
                "26634432",
                "protein_xref_glygen_ds",
                "GLY_000492",
                "",
                ""
            ],
            [
                "O55188-1",
                "Glycosylation of dentin matrix protein 1 is a novel key element for astrocyte maturation and BBB integrity.",
                "Protein & cell",
                "2018 Mar",
                "Jing B, Zhang C, Liu X, Zhou L, Liu J, Yao Y, Yu J, Weng Y, Pan M, Liu J, Wang Z, Sun Y, Sun YE",
                "protein_xref_pubmed",
                "28822114",
                "protein_xref_glygen_ds",
                "GLY_000492",
                "",
                ""
            ]
        ]
    },
    "readme": "Name\n   - Mouse Proteoform Citations (Literature Mining)\n\nVersion\n   - v-2.12.1\n\nCreated on\n  - 2021-06-04T10:20:39.827303\n\nModified on\n   - 2025-04-15T14:29:15.805Z\n\nCreated by\n   - Jeet Vora [jeetvora@gwu.edu], Robel Kahsay [rykahsay@gwu.edu], Karina Martinez [karinamartinez@gwu.edu], Luke Johnson [luke.johnson@email.gwu.edu], Urnisha Bhuiyan [sbhuiyan28@email.gwu.edu], Cyrus Au-Yeung [cyruschauyeung@gwu.edu], Kate Warner [k.warner1@gwu.edu]\n\nAuthors\n   - \n\nChecksum\n   - \n\nUsability Domain\n   - The Mouse Proteoform Citations [Automatic Literature Mining] dataset contains Mus musculus [taxid:10090] citation details of the publications present in the Mouse Glycosylation Sites [Automatic Literature Mining] [Automatically verified] dataset (https://data.glygen.org/GLY_000492).\n\nDescription Domain\n   Keywords\n   - PMID\n   - journal\n   - citations\n   - authors\n   - Literature\n   - PubMed\n\n   Pipeline Steps\n   - Step-1: The script parses the input file for PMIDs and accesses the NCBI Entrez infrastructure to download the publication information of the corresponding PMIDs.\n\nExecution Domain\n   Scripts   - https://github.com/glygener/glygen-backend-integration/blob/master/dataset-maker/make-proteoform-dataset.py\n\n\n   Software Prerequisites\n      Name: Python\n      Version: 2.7.5\n      URI: https://www.python.org/download/releases/2.7.5/\n\nI/O Domain\n   Input Subdomain\n      Name: all_blacklisted_pmids.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/compiled/all_blacklisted_pmids.csv\n   Output Subdomain\n      Name: mouse_proteoform_citations_glycosylation_sites_literature_mining.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/mouse_proteoform_citations_glycosylation_sites_literature_mining.csv\n      Name: mouse_proteoform_citations_glycosylation_sites_literature_mining.stat.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/mouse_proteoform_citations_glycosylation_sites_literature_mining.stat.csv\n\nFields\n      Name: \"uniprotkb_canonical_ac\"\"\n     Description: \"\"\n      Unique Values: \"69\"",
    "downloadurl": "/ln2data/releases/data/v-2.12.1/reviewed/mouse_proteoform_citations_glycosylation_sites_literature_mining.csv"
}