wateRmelon (version 1.16.0)

dasen-methods: Calculate normalized betas from MethyLumiSets of Illumina 450K methylation arrays

Description

Multiple ways of calculating the index of methylation (beta) from methylated and unmethylated probe intensities used in Pidsley et al 2012.

Arguments

mn, mns, data, bn

A MethyLumiSet object. Sample names names are used to get Sentrix row and column by default, see '...'.

fudge

value added to total intensity to prevent denominators close to zero when calculating betas

...

additional argument roco for dfsfit giving Sentrix rows and columns. This allows a background gradient model to be fit. This is split from data column names by default. roco=NULL disables model fitting (and speeds up processing), otherwise roco can be supplied as a character vector of strings like 'R01C01' (only 3rd and 6th characters used).

Value

a matrix (default method) or object of the same shape and order as the first argument containing betas.

methods

dasen ( mns, fudge = 100, ... ) nasen ( mns, fudge = 100 ) betaqn( bn ) naten ( mn, fudge = 100 ) naten ( mn, fudge = 100 ) nanet ( mn, fudge = 100 ) nanes ( mns,fudge = 100 ) danes ( mn, fudge = 100, ... ) danet ( mn, fudge = 100, ... ) danen ( mns,fudge = 100, ... ) daten1( mn, fudge = 100, ... ) daten2( mn, fudge = 100, ... ) tost ( mn ) fuks ( data) swan ( mn )

Details

dasen same as nasen but type I and type II backgrounds are normalized first. This is our recommended method

betaqn quantile normalizes betas

naten quantile normalizes methylated and unmethylated intensities separately, then calculates betas

nanet quantile normalizes methylated and unmethylated intensities together, then calculates betas. This should equalize dye bias.

nanes quantile normalizes methylated and unmethylated intensities separately, except for type II probes where methylated and unmethylated are normalized together. This should equalize dye bias without affecting type I probes which are not susceptible.

danes same as nanes, except typeI and type II background are equalised first.

danet same as nanet, except typeI and type II background are equalised first.

danen background equalisation only, no normalization

daten1 same as naten, except typeI and type II background are equalised first (smoothed only for methylated)

daten2 same as naten, except typeI and type II background are equalised first (smoothed for methylated an unmethylated)

nasen same as naten but typeI and typeII intensities quantile normalized separately

tost method from Touleimat and Tost 2011

fuks method from Dedeurwaerder et al 2011. Peak correction only, no normalization

swan method from Maksimovic et al 2012

References

[1] Pidsley R, Wong CCY, Volta M, Lunnon K, Mill J, Schalkwyk LC: A data-driven approach to preprocessing Illumina 450K methylation array data (submitted)

[2] Dedeurwaerder S, Defrance M, Calonne E, Sotiriou C, Fuks F: Evaluation of the Infinium Methylation 450K technology . Epigenetics 2011, 3(6):771-784.

[3] Touleimat N, Tost J: Complete pipeline for Infinium R Human Methylation 450K BeadChip data processing using subset quantile normalization for accurate DNA methylation estimation. Epigenomics 2012, 4:325-341

[4] Maksimovic J, Gordon L, Oshlack A: SWAN: Subset quantile Within-Array Normalization for Illumina Infinium HumanMethylation450 BeadChips. Genome biology 2012, 13(6):R44