Snakefile

include: "../common.smk"
configfile: "config.yml"

# https://doi.org/10.1016/j.isci.2022.104097

# cellxgene portal AWS URLs expire in one week.
# May need to get new URLs from https://cellxgene.cziscience.com/collections/8e880741-bf9a-4c8e-9227-934204631d2a

# High Resolution Slide-seqV2 Spatial Transcriptomics Enables Discovery of Disease-Specific Cell Neighborhoods and Pathways

# Spatial transcriptomics in the healthy human kidney: Puck_200903_13
H5AD_URL = "https://corpora-data-prod.s3.amazonaws.com/c4bcefdb-6c4a-415e-95a4-eb73a41c734f/local.h5ad?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=ASIATLYQ5N5XQI3KD7NP%2F20221212%2Fus-west-2%2Fs3%2Faws4_request&X-Amz-Date=20221212T164848Z&X-Amz-Expires=604800&X-Amz-SignedHeaders=host&X-Amz-Security-Token=IQoJb3JpZ2luX2VjEH0aCXVzLXdlc3QtMiJHMEUCIDrHa8FXitpViXVHw%2B9i2hmHNyPFznFdjvTQHoU372OsAiEAs4fAZ%2B%2BPWLoQlyx8o9Wv8puh0IkZ7xjXroY3G4oODAAq9AMIpv%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FARABGgwyMzE0MjY4NDY1NzUiDNlfVhvq4gVG5tH6hSrIAwsDF3CZCFVhMGxIRXjsKW0uyWGqttGfNYLiFEqc4Xfi1MSN%2BD2Xv1Q1HYQhQ4qkb7xdpTORVV6XZkaUUafAXuhNaMZwsXOM%2FEvpytZdyLXAPHHvundl6eKAOwUplPVGIsC8gJnGKpJYhvmpkpSctoY%2B0eje5Ri5xCKRajMpZSJykMYqVkTMOzOb1WE38sLJoHk9P%2Bu1aarAFJQPCQMqwVj7wjl54RnUtetrpp5VZsrya2mMQAUmflyXI5hW39wYnLgbl%2F%2B6Nf80G7h50d2YHcMVGwZ2lCV%2B5zjQ5cE6iuMym9zrBR062M7%2Fkc3UnQ5Eulzp28CFhfsKUdal7t89btJIQo6X9Z9X6kG7MTp%2BoNaAkDDD2Idcu6itdeFkNSm%2BlGKCwchzgiAYz%2FG51HS30RlHYZSrSJhRP0Q4UIt%2FlGTnwwpeA5ea5aK5YM4xUNBq16XD1sN8X47d5ZwK2HvpAgPCaqLqe2R4Fi%2FK0lns5%2F%2BmQK4u6Wsnp%2F5KjAkyQK3k3LcM9g5xvqn5udFWp4BAMRgnlw8PwaffrMA8te%2BMryhsqyIv757%2BSr%2FgdncfLuIR%2FgMRqXhQM%2BF0aaiAig%2FsL3X5fKXmQI6AVjC%2B0dycBjqlAZsOF%2F8GZzh54w8egq2UEczooQF6PzpmN946hUSjvSF1Gje7o4mhagBeaS56JTI3AYBsHvmpW%2Fl0vFUcYLJJsWCHCOxQQsYLNiPSOowpMLJNchGGpg882LVCOBGjGhqi5jrJHyYRFQXTYPStgMXDfD%2Fu70GqRjTSd5DKENx1xWHbUSyObOebz90ZnUZmd5sHKtkCtMCLsjBqepWTFOYqtoSFFKDGMw%3D%3D&X-Amz-Signature=aa7763df7667dc638aec949a1279e28eea9972d1a64a228f85e6224a457dd95b"

rule all:
    input:
        [ (PROCESSED_DIR / f) for f in config['output'] ]

rule convert_to_zarr:
    input:
        (RAW_DIR / "marshall_2022_iscience.h5ad")
    output:
        directory(PROCESSED_DIR / "marshall_2022_iscience.h5ad.zarr")
    params:
        script=(SRC_DIR / "convert_to_zarr.py")
    shell:
        '''
        python {params.script} \
            -i {input} \
            -o {output}
        '''

# Download raw h5ad file.
rule download_adata:
    output:
        (RAW_DIR / "marshall_2022_iscience.h5ad")
    params:
        file_url=H5AD_URL
    shell:
        '''
        curl -L -o {output} "{params.file_url}"
        '''