prior¶

Purpose¶

Computes base priors per candidate locus using gnomAD reference data.

Upstream¶

• umi_combine

Required inputs¶

• gnomad_path
• filter_mpileup_file
• genome autosome metadata from genome_details

Input interpretation¶

Input key	Source	Required	Interpretation
gnomad_path	resource_details.gnomad_path	Yes	Population AF resource directory/list used for prior lookup.
filter_mpileup_file	mpileup output	Yes	Candidate-locus file used as prior-query target list.
genome_details	config loader	Yes	Chromosome metadata used for per-chromosome resource querying.

Parameters¶

No major step-level tunables are exposed in config; this step uses query-driven prior lookup and normalizes per-base priors (A/T/C/G) per site.

Parameter interpretation¶

Parameter area	Interpretation
gnomad_path content/coverage	Primary determinant of prior completeness and quality.
chromosome metadata (genome_details)	Controls which chromosomes are queried and merged into prior output.
run.threads (indirect)	Affects overall pipeline throughput; this step itself is mostly I/O/resource-lookup driven.

Outputs¶

• prior_file: output_dir/prior.txt

Tuning notes¶

• gnomAD input can be either a directory or a file list.
• Missing chromosome resources are tolerated, but reduce prior completeness.