Skip to contents

Selects denominator anchors with the lowest ratio of between-biofluid variance to total variance in the training pool.

Usage

fit_biofluid_anchor_rclr(
  train_expr,
  train_meta,
  panel,
  biofluid_col = "biofluid",
  pseudocount = NULL,
  anchor_n = 12L,
  min_anchor_n = 3L,
  min_biofluid_classes = 2L,
  exclude_features = .bfa_default_blood_cell_mirnas()
)

Arguments

train_expr

Numeric matrix, samples x features.

train_meta

Data frame aligned to `train_expr`.

panel

Character vector of panel features.

biofluid_col

Column in `train_meta` with canonical or raw biofluid.

pseudocount

Optional additive pseudocount before log transform.

anchor_n

Number of anchor features to select.

min_anchor_n

Minimum anchors required for an evaluable fit.

min_biofluid_classes

Minimum biofluid classes in which an anchor must be detected.

exclude_features

Features excluded from rCLR denominator selection.