mecfs_bio.assets.gwas.csf_pqtl.raw.csf_template_aptamer

The template aptamer for the CSF pQTL variant index: one arbitrary aptamer's GWAS-SSF summary statistics, downloaded and converted to parquet.

ConstructCsfVariantIndexTask is templated off a single aptamer's variant set (the CSF PLINK2 run is joint across aptamers, so one aptamer is representative). This plays the role STACK_UKBBPPP_RABGAP1L plays for the PPP index.

The download is the raw .tsv.gz; PipeDataFrameTask reads the whole 7.3M-row file with the polars backend (required for text sources) and rewrites it as parquet so the index task can scan it cheaply. The read_spec on the download meta declares the tab separator PipeDataFrameTask reads through.

Attributes:

• CSF_TEMPLATE_APTAMER –
• CSF_TEMPLATE_APTAMER_DOWNLOAD –

CSF_TEMPLATE_APTAMER module-attribute

CSF_TEMPLATE_APTAMER = PipeDataFrameTask.create(
    source_task=CSF_TEMPLATE_APTAMER_DOWNLOAD,
    asset_id="csf_template_aptamer_parquet",
    out_format=ParquetOutFormat(),
    pipes=[],
    backend="polars",
)

CSF_TEMPLATE_APTAMER_DOWNLOAD module-attribute

CSF_TEMPLATE_APTAMER_DOWNLOAD = DownloadFileTask(
    meta=ReferenceFileMeta(
        group="csf_pqtl",
        sub_group="western_2024",
        sub_folder=PurePath("raw"),
        id=AssetId("csf_template_aptamer_sumstats"),
        filename="csf_template_aptamer",
        extension=".tsv.gz",
        read_spec=DataFrameReadSpec(
            DataFrameTextFormat(separator="\t")
        ),
    ),
    url=gwas_catalog_sumstats_url(_TEMPLATE_ACCESSION),
    md5_hash=_TEMPLATE_MD5,
)