mt_subset: Normalise and subset the data in 'mmt' objects based on...

In TYMichaelsen/mmtravis: Tools for analysing (meta)transcriptomics data

Description

Subset mt objects based on sample metadata

Usage

mt_subset(data, ...)

Arguments

mmt	(required) Data list as loaded with mt_load.
sub_genes	(optional) A string specifying a logical row subset operation on the mtgene dataframe in the mt object parsed to subset.
sub_samples	(optional) A string specifying a logical row subset operation on the mtmeta dataframe in the mt object parsed to subset.
minreads	Minimum number of reads pr. gene. Genes below this value will be removed. (default: 0)
frac0	Fraction of zeros allowed per gene. Genes with a higher fraction of zeros will be removed. (default: 1)
normalise	Normalise the read counts AFTER reads have been removed by the minreads argument but BEFORE any sample/gene subsetting. (default: "none") "quantile": Quantile normalization. See normalize.quantiles for details. "TPM": Normalise read counts to Transcripts Per Milion (TPM). "abundance": Normalise read counts to relative abundance. "libsize": Normalise the read counts to adjust for gene dispersion and total read counts per sample. See estimateSizeFactorsForMatrix for details. "vst": Normalise as "libsize" and perform robust log2-transformation. See vst for details. "log2": Perform log2(x + 1)-transformation. "none": No normalisation.

Details

The function mt_subset operates in three steps:

• Filter - genes according to minreads and frac0.

• Normalise - as specified in normalise.

• Subset - according to sub_genes and sub_samples.

Subsetting are performed on the mtmeta/mtgene data by subset and the whole object is then adjusted accordingly.

Value

A modifed mt object

Author(s)

Thomas Yssing Michaelsen [e-mail]

Examples

## Not run: 
# Get some data.
data("example_mmt")

# Let's subset to contig 1, 7675, and 69676.
mt1 <- mt_subset(example_mmt,sub_genes = "contig %in% c(1,7675,69676)")
mt1

# Let's subset to specific organism.
mt2 <- mt_subset(example_mmt,sub_samples = "Organism == 'Brocadia'")
mt2

# Let's do both and remove genes with less than 10000 reads in total.
mt3 <- mt_subset(example_mmt,
  sub_samples = "Organism == 'Brocadia'",
  sub_genes   = "contig %in% c(1,7675,69676)",
  minreads    = 10000)
mt3

# You can also normalise the data and subset.
mt4 <- mt_subset(example_mmt,
  sub_samples = "Organism == 'Brocadia'",
  sub_genes   = "contig %in% c(1,7675,69676)",
  minreads    = 10000,
  normalise   = "libsize")
mt4 # Note "Normalised:" is now included.

## End(Not run)

TYMichaelsen/mmtravis documentation built on Aug. 5, 2019, 10:48 a.m.