{
    "bcoid": "GLY_001268",
    "title": "Zebrafish Protein Citations (UniProtKB)",
    "description": "The Zebrafish Protein Citations dataset contains publication information for Danio rerio [taxid:7955] ...",
    "filename_list": "zebrafish_protein_citations_uniprotkb.csv",
    "filename": "zebrafish_protein_citations_uniprotkb.csv",
    "filetype": "csv",
    "categories": {
        "species": [
            "Danio rerio"
        ],
        "molecule": [
            "Protein"
        ],
        "tag": [
            "Protein Citations"
        ],
        "file_type": [
            "csv"
        ],
        "status": [
            "Reviewed"
        ],
        "scope": [
            "internal"
        ],
        "file_category": [
            "citations"
        ]
    },
    "minitable": {
        "content": [
            [
                "Q6DGB6-1...",
                "CCDC103 muta..."
            ],
            [
                "A0A2R8QFQ6-1...",
                "De Novo Miss..."
            ]
        ],
        "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"
                }
            ],
            [
                "Q6DGB6-1",
                "CCDC103 mutations cause primary ciliary dyskinesia by disrupting assembly of ciliary dynein arms.",
                "Nature genetics",
                "2012 May 13",
                "Panizzi JR, Becker-Heck A, Castleman VH, Al-Mutairi DA, Liu Y, Loges NT, Pathak N, Austin-Tse C, Sheridan E, Schmidts M, Olbrich H, Werner C, Haffner K, Hellman N, Chodhari R, Gupta A, Kramer-Zucker A, Olale F, Burdine RD, Schier AF, O'Callaghan C, Chung EM, Reinhardt R, Mitchison HM, King SM, Omran H, Drummond IA",
                "protein_xref_pubmed",
                "22581229",
                "protein_xref_uniprotkb",
                "Q6DGB6"
            ],
            [
                "A0A2R8QFQ6-1",
                "De Novo Missense Variants in FBXW11 Cause Diverse Developmental Phenotypes Including Brain, Eye, and Digit Anomalies.",
                "American journal of human genetics",
                "2019 Sep 5",
                "Holt RJ, Young RM, Crespo B, Ceroni F, Curry CJ, Bellacchio E, Bax DA, Ciolfi A, Simon M, Fagerberg CR, van Binsbergen E, De Luca A, Memo L, Dobyns WB, Mohammed AA, Clokie SJH, Zazo Seco C, Jiang YH, Sorensen KP, Andersen H, Sullivan J, Powis Z, Chassevent A, Smith-Hicks C, Petrovski S, Antoniadi T, Shashi V, Gelb BD, Wilson SW, Gerrelli D, Tartaglia M, Chassaing N, Calvas P, Ragge NK",
                "protein_xref_pubmed",
                "31402090",
                "protein_xref_uniprotkb",
                "A0A2R8QFQ6"
            ]
        ]
    },
    "readme": "Name\n   - Zebrafish Protein Citations (UniProtKB)\n\nVersion\n   - v-2.12.1\n\nCreated on\n  - \n\nModified on\n   - \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   - Maria Martin [martin@ebi.ac.uk], Preethi Vasudev [pvasudev@ebi.ac.uk]\n\nChecksum\n   - \n\nUsability Domain\n   - The Zebrafish Protein Citations dataset contains publication information for Danio rerio [taxid:7955] UniProtKB accessions that includes PMID, journal name, date and author and title. Publications of large scale studies, and high frequency publications pertaining to genome sequencing, cDNA, chromosomes have been filtered out. If you use this dataset please provide proper attribution to EMBL-EBI-UniProtKB and GlyGen.\n\nDescription Domain\n   Keywords\n   - publications\n   - PMID\n   - journal\n   - citations\n   - authors\n   - UniProtKB\n   - PubMed\n\n   Pipeline Steps\n   - Step-1: The script loads the UniProtKB publication data from the triplestore containing triples from nt file to retrieve the publication information associated with the UniProtKB accession and maps it to the UniProtKB canonical accessions. The script also filters out publications of large scale studies, and high frequency publications pertaining to DNA sequencing, cDNA, chromosomes using blacklisted PMIDs file.\n\nExecution Domain\n   Scripts   - https://github.com/glygener/glygen-backend-integration/blob/master/dataset-maker/make-protein-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      Name: zebrafish_protein_function_uniprotkb.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/zebrafish_protein_function_uniprotkb.csv\n      Name: zebrafish_protein_ptm_annotation_uniprotkb.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/zebrafish_protein_ptm_annotation_uniprotkb.csv\n      Name: zebrafish_protein_site_annotation_uniprotkb.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/zebrafish_protein_site_annotation_uniprotkb.csv\n      Name: uniprot-proteome-danio-rerio.nt\n      URI: https:/data.glygen.org/ln2downloads/ebi/2026_06_13/uniprot-proteome-danio-rerio.nt\n   Output Subdomain\n      Name: zebrafish_protein_citations_uniprotkb.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/zebrafish_protein_citations_uniprotkb.csv\n      Name: zebrafish_protein_citations_uniprotkb.stat.csv\n      URI: https:/data.glygen.org/ln2releases/data/v-2.12.1/reviewed/zebrafish_protein_citations_uniprotkb.stat.csv\n\nFields\n      Name: \"uniprotkb_canonical_ac\"\"\n     Description: \"\"\n      Unique Values: \"7293\"",
    "downloadurl": "/ln2data/releases/data/v-2.12.1/reviewed/zebrafish_protein_citations_uniprotkb.csv"
}