normalize.batches: Batch normalize the X & Y intensity data.

In dmgatti/DOQTL: Genotyping and QTL Mapping in DO Mice

Batch normalize the X & Y intensity data.

Description

This function batch normalizes the X & Y intensity data by subtracting batch medians from the X & Y intensities.

Usage

  batch.normalize(path = ".", snps)
  quantilenorm(x1, y1, x2, y2)

Arguments

path	Character, the full path to the input files, which must be either "x.txt" and "y.txt" or "x.filt.txt" and "y.filt.txt".
snps	Data.frame, with three columns containing SNP ID, chromosome and Mb location in that order. May be obtained from ftp://ftp.jax.org/MUGA.
x1	Numeric matrix containing X intensities for batch 1 containing samples in rows and markers in columns. Number of samples should be larger than x2.
y1	Numeric matrix containing Y intensities for batch 1 containing samples in rows and markers in columns. Number of samples should be larger than y2.
x2	Numeric matrix containing X intensities for batch 2 containing samples in rows and markers in columns.
y2	Numeric matrix containing Y intensities for batch 2 containing samples in rows and markers in columns.

Details

quantile.norm adjusts the intensities of samples in batch 2 to those of batch 1. The number of samples in batch 1 should be greater than the number of samples in batch 2. At each SNP, we form quantiles of the X1 (or Y1) intensity distribution, discarding the upper and lower 0.01

Value

FALSEor batch.normalize: returns value is returned. The batch normalized intensities are written to "x.filt.batch.norm.txt" and "y.filt.batch.norm.txt".

FALSEor quantilenorm: returns normalized X and Y values for batch 2.

Note

FALSEuture releases may include more sophisticated normalization algorithms.

Author(s)

Daniel Gatti

See Also

extract.raw.data, filter.samples

Examples

  ## Not run: 
    load(url("ftp://ftp.jax.org.MUGA/muga_snps.Rdata"))
    batch.normalize(path = "/demo/MUGA/", snps = muga_snps)
  
## End(Not run)

dmgatti/DOQTL documentation built on April 7, 2024, 10:35 p.m.