Package 'PONG2'

Train KIR prediction models in parallel

Description

Train KIR genotype prediction models using parallel attribute bagging across multiple CPU cores. This is the core training function used by the pong2 train CLI command.

Usage

kirParallelAttrBagging(
  cl,
  hla,
  snp,
  auto.save = "",
  nclassifier = 100,
  mtry = c("sqrt", "all", "one"),
  prune = TRUE,
  rm.na = TRUE,
  stop.cluster = FALSE,
  verbose = TRUE
)

Arguments

cl	a cluster object created by parallel::makeCluster() for parallel computation across multiple CPU cores
hla	a KIR allele table object of class hlaAlleleClass containing training allele calls
snp	a SNP genotype object of class hlaSNPGenoClass containing training SNP data
auto.save	character string; file path prefix for auto-saving classifiers during training. Use "" (default) to disable
nclassifier	integer; number of individual ensemble classifiers to train (default: 100)
mtry	character; number of SNPs randomly selected at each node. One of "sqrt" (default), "all", or "one"
prune	logical; if TRUE (default), prune classifiers
rm.na	logical; if TRUE (default), remove samples with missing KIR allele calls
stop.cluster	logical; if TRUE, stop the parallel cluster after training (default: FALSE)
verbose	logical; if TRUE (default), print progress

Value

An object of class hlaAttrBagClass representing the trained PONG2 KIR prediction model. The object contains:

n.samp

integer; number of training samples

n.snp

integer; number of SNP predictors used

hla.locus

character; the KIR locus name

hla.allele

character vector; KIR alleles in the model

classifiers

list; individual ensemble classifiers

out.of.bag.acc

numeric; out-of-bag accuracy estimate

Use kirPredict() to apply the model to new samples.

Examples

# Load example data
data(PONG2_example)

# Set up parallel cluster
cl <- parallel::makeCluster(2)

# Train a small model
model <- kirParallelAttrBagging(
  cl          = cl,
  hla         = example_kir,
  snp         = example_snp,
  nclassifier = 20,
  verbose     = FALSE
)

parallel::stopCluster(cl)

# View model summary
print(model)

# Clean up
hlaClose(model)

---

Predict KIR genotypes from SNP data

Description

Predict KIR genotypes for a set of samples using a trained PONG2 attribute bagging model. This is the core prediction function used by the pong2 impute CLI command.

Usage

kirPredict(
  object,
  snp,
  cl = FALSE,
  type = c("response+dosage", "response", "prob", "response+prob"),
  vote = c("prob", "majority"),
  allele.check = TRUE,
  match.type = c("Position", "Pos+Allele", "RefSNP+Position", "RefSNP"),
  same.strand = FALSE,
  verbose = TRUE,
  verbose.match = TRUE
)

Arguments

object	a PONG2 model object of class hlaAttrBagClass as returned by kirParallelAttrBagging()
snp	a SNP genotype object of class hlaSNPGenoClass containing the target samples to impute
cl	a cluster object for parallel computation, or FALSE (default) for single-threaded prediction
type	character; type of prediction output. One of: "response+dosage" (default; predicted alleles + dosage scores), "response" (predicted alleles only), "prob" (posterior probabilities only), "response+prob" (predicted alleles + posterior probabilities)
vote	character; voting method for ensemble classifiers. One of "prob" (default; probability-weighted voting) or "majority" (majority vote)
allele.check	logical; if TRUE (default), check and validate allele names against the model
match.type	character; SNP matching method. One of "Position" (default), "Pos+Allele", "RefSNP+Position", or "RefSNP"
same.strand	logical; if TRUE, assume SNPs are on the same strand (default: FALSE)
verbose	logical; if TRUE (default), print progress messages
verbose.match	logical; if TRUE (default), print SNP matching summary

Value

An object of class hlaAlleleClass containing KIR imputation results. The object includes:

value

data frame with columns sample.id, allele1, allele2, and prob (posterior probability of the best call)

dosage

numeric matrix of allele dosage scores (samples x alleles); NULL if type = "response"

postprob

numeric matrix of posterior probabilities (alleles x samples); NULL unless type = "response+prob" or "prob"

Samples with posterior probability below the call threshold (CT) are assigned NA for both alleles.

Examples

# Load example data
data(PONG2_example)

# Load model from object
model <- hlaModelFromObj(example_mobj)

# Predict KIR genotypes
pred <- kirPredict(
  object  = model,
  snp     = example_snp,
  type    = "response+prob",
  verbose = FALSE
)

# View results
head(pred$value)

# Clean up
hlaClose(model)

---

PONG2 Example Dataset

Description

A small example dataset for demonstrating PONG2 functions. Contains 50 samples and 200 SNPs in the KIR region (chr19), along with a pre-trained KIR3DL1 model with 10 classifiers.

Usage

data(PONG2_example)

Format

Three objects are loaded:

example_snp

A hlaSNPGenoClass object with 50 samples and 200 SNPs in the KIR region (chr19, hg19 assembly)

example_kir

A hlaAlleleClass object with KIR3DL1 allele calls for 50 samples

example_mobj

A hlaAttrBagObj object — a pre-trained KIR3DL1 model with 10 ensemble classifiers

Examples

data(PONG2_example)
# SNP data
cat("Samples:", ncol(example_snp$genotype), "\n")
cat("SNPs:   ", nrow(example_snp$genotype), "\n")
# KIR allele table
cat("Locus:  ", example_kir$locus, "\n")
# Model
cat("Classifiers:", length(example_mobj$classifiers), "\n")

---