GT_MPI_GATHER(1)

💖 GT_MPI_GATHER 💖

okay bestie, gather round (literally, it's in the name) — this is gt_mpi_gather, the MPI gatherer for GenomicsDB. yes it lives in section 1 of the manual, and yes it's about to become your whole personality. 🧚✨

🌸 NAME 🌸

gt_mpi_gather — the MPI gatherer for GenomicsDB, no cap 💅

✨ SYNOPSIS ✨

gt_mpi_gather [options]

that's it. that's the whole vibe. you give it options, it gives you variants. periodt. 👑

🎀 OPTIONS 🎀

• --help, -h — prints a usage message that summarizes all the options you've got available, then dips out. your little safety net, bestie. 🫶
• --json-config=<query json file>, -j <query json file> — this is the main character right here. you can specify workspace, array, query_column_ranges, query_row_ranges, vid_mapping_file, callset_mapping_file, query_attributes, query_filter, reference_genome, etc. as fields inside the json file. like, for example:

{ "workspace" : "/tmp/ws",
       "array" : "t0_1_2", "query_column_ranges" : [ [ [0, 100 ], 500 ]
       ], "query_row_ranges" : [ [ [0, 2 ] ], "vid_mapping_file" :
       "/tests/inputs/vid.json", "callset_mapping_file":
       "/tests/inputs/callset_mapping.json", "query_attributes" : [
       "REF", "ALT", "BaseQRankSum", "MQ", "MQ0", "ClippingRankSum",
       "MQRankSum", "ReadPosRankSum", "DP", "GT", "GQ", "SB", "AD",
       "PL", "DP_FORMAT", "MIN_DP" ] }

• --loader-json-config=<loader json file>, -l <loader json file> — this one is optional, but only if the vid_mapping_file and callset_mapping_file fields are already specified in the query json file. so like, if your query json already spilled the tea, you don't need this. 🙈
• --workspace=<workspace dir>, -w <GenomicsDB workspace dir> — optional, and only if workspace is specified in any of the json config files. iykyk. 💜
• --array=<array dir>, -A <GenomicsDB array dir> — also optional, and only if array is specified in any of the json config files. same energy as above, different field. 🦋
• --print-calls — optional! prints VariantCalls in a JSON format. slay. ✨
• --print-csv — optional! outputs CSV, where the fields and the order of the CSV lines are determined by the query attributes. organized queen. 📊
• --produce-Broad-GVCF — optional! produces a combined gVCF from the GenomicsDB data, constrained by the query configuration. iconic. 🔥
• --output-format=<output_format>, -O <output_format> — used with --produce-Broad-GVCF. the output format can be one of these strings: "z[0-9]" (that's compressed VCF), "b[0-9]" (compressed BCF) or "bu" (uncompressed BCF). the default is uncompressed VCF if you don't specify anything. 💅
• --produce-histogram — optional. that's the whole description bestie, it's giving mystery. 🌈
• --produce-interesting-positions — optional. also just vibes. ⭐
• --version — prints the version and exits. quick little flex. 👑

fr fr, if you don't specify any of the print/produce arguments, the tool prints all the Variants constrained by the query configuration in a JSON format. no arguments, no problem. 😌

🌟 Parallel Querying 🌟

MPI can be used for parallel querying, bestie — like, actually parallel. for example:

mpirun -n <num_processes> -hostfile <hostfile> ./bin/gt_mpi_gather -j <query.json> -l <loader.json> [<other_args>]

look at all those processes doing their thing at the same time. main character energy, truly. 🫶💖

and that's the whole page, gorgeous. this was gt_mpi_gather from GT_MPI_GATHER(1), hanging out in July 2022 in the User Commands section. go gather some variants, you absolute legend. 🎀✨💅