Skip to contents

Derive SSD parameters for msPAF computation

Source: R/mspaf.R
derive_ssd_params.Rd

Reads analyte eligibility and mode-of-action groups from the bundled metadata CSV, then calls ssd_hc50() and .ssd_sigma() to populate the HC50 and effective sigma needed for Concentration Addition. Chemistry normalisation formulas (currently all stubs) are parsed once and stored as a list column for use in compute_mspaf_per_sample().

Usage

derive_ssd_params(meta, method, guideline_dir)

Arguments

meta

Analyte metadata tibble from .load_analyte_metadata().

method

SSD method: "multi" or "anzecc".

guideline_dir

Path to ANZG guideline data folder.

Value

Tibble with columns analyte, hc50, sigma, moa_group, parsed_formula (list of language objects or NULLs), coanalytes_req (character), and fit (list column of fitted SSD objects, one per analyte). The fitted SSD is loaded once per analyte here (via the cached .load_or_fit() path) so that compute_mspaf_per_sample() can evaluate every sample's PAF in a single vectorised ssdtools::ssd_hp() call per analyte, rather than refitting / re-resolving the SSD inside a per-(sample × analyte) loop.

Details

Eligibility criteria:

  • ssd_available == TRUE in the metadata

  • analyte not in .MSPAF_EXCLUDED_ANALYTES

  • ssd_hc50() returns a non-NA value (model fits successfully)