CrossPlatformAdapter R6 Class

An adapter for learning a source-platform representation and applying it to target-platform data before model transfer.

Value

A `CrossPlatformAdapter` object.

Methods

Public methods

• CrossPlatformAdapter$new()

• CrossPlatformAdapter$fit()

• CrossPlatformAdapter$adapt()

• CrossPlatformAdapter$get_source_data()

• CrossPlatformAdapter$is_fitted()

• CrossPlatformAdapter$print()

• CrossPlatformAdapter$clone()

---

Method new()

Create a cross-platform adapter with the selected normalization strategy and backend options.

Usage

CrossPlatformAdapter$new(
  strategy = c("rank", "quantile", "zscore", "reference", "combat_pooled"),
  reference_features = NULL,
  reference_summary = c("mean", "median"),
  source_platform = "source",
  target_platform = "target",
  combat_parametric = TRUE,
  combat_mean_only = FALSE
)

Arguments

strategy

Adaptation strategy. One of `"rank"`, `"quantile"`, `"zscore"`, `"reference"`, or `"combat_pooled"`.

reference_features

Optional character vector of reference miRNAs used when `strategy = "reference"`.

reference_summary

Summary statistic for reference normalization: `"mean"` or `"median"`.

source_platform

Label used for source samples in pooled ComBat.

target_platform

Label used for target samples in pooled ComBat.

combat_parametric

Logical, use parametric priors in pooled ComBat.

combat_mean_only

Logical, only correct means in pooled ComBat.

---

Method fit()

Learn a source-platform reference representation.

Usage

CrossPlatformAdapter$fit(source_data, source_labels = NULL)

Arguments

source_data

Numeric matrix with samples in rows and features in columns.

source_labels

Optional labels for the source data. Required for `strategy = "combat_pooled"`.

Returns

The adapter itself, invisibly.

---

Method adapt()

Adapt target-platform data to the learned source representation.

Usage

CrossPlatformAdapter$adapt(target_data, target_labels = NULL)

Arguments

target_data

Numeric matrix with samples in rows and features in columns.

target_labels

Optional target labels. Required when `strategy = "combat_pooled"`.

Returns

Adapted target matrix with samples in rows and features in columns.

---

Method get_source_data()

Return the stored source matrix.

Usage

CrossPlatformAdapter$get_source_data(adapted = TRUE)

Arguments

adapted

Logical, return the adapted source matrix if `TRUE`, otherwise return the raw source matrix.

Returns

A numeric matrix.

---

Method is_fitted()

Check whether the adapter has been fitted.

Usage

CrossPlatformAdapter$is_fitted()

Returns

Logical scalar.

---

Method print()

Print a compact adapter summary.

Usage

CrossPlatformAdapter$print()

---

Method clone()

The objects of this class are cloneable with this method.

Usage

CrossPlatformAdapter$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

set.seed(42)
src <- matrix(rnorm(100 * 5), 100, 5)
tgt <- matrix(rnorm(50 * 5, mean = 2), 50, 5)

adapter <- CrossPlatformAdapter$new(strategy = "rank")
adapter$fit(src)
adapted_tgt <- adapter$adapt(tgt)
head(adapted_tgt)
#>      feature_1 feature_2 feature_3 feature_4 feature_5
#> [1,]      1.00      0.00      0.25      0.75      0.50
#> [2,]      1.00      0.25      0.75      0.00      0.50
#> [3,]      0.75      0.50      1.00      0.00      0.25
#> [4,]      0.25      1.00      0.75      0.50      0.00
#> [5,]      0.25      0.75      0.50      1.00      0.00
#> [6,]      0.50      0.75      1.00      0.00      0.25