Estimate divergence against a given reference sample.
addDivergence(x, name = "divergence", ...)
# S4 method for class 'SummarizedExperiment'
addDivergence(x, name = "divergence", ...)
getDivergence(
x,
assay.type = assay_name,
assay_name = "counts",
reference = "median",
method = "bray",
...
)
# S4 method for class 'SummarizedExperiment'
getDivergence(
x,
assay.type = assay_name,
assay_name = "counts",
reference = "median",
method = "bray",
...
)a SummarizedExperiment object.
Character scalar. The name to be used to store the result
in metadata of the output. (Default: method)
optional arguments passed to
addDissimilarity. Additionally:
dimred: Character scalar. Specifies the name of
dimension reduction result from reducedDim(x). If used, these
values are used to calculate divergence instead of the assay. Can be
disabled with NULL. (Default: NULL)
Character scalar. Specifies which assay to use for
calculation. (Default: "counts")
Deprecated. Use assay.type instead.
Character scalar. A column name from
colData(x) or either "mean" or "median".
(Default: "median")
Character scalar. Specifies which dissimilarity to
calculate. (Default: "bray")
x with additional colData named name
Microbiota divergence (heterogeneity / spread) within a given sample set can be quantified by the average sample dissimilarity or beta diversity with respect to a given reference sample.
The calculation makes use of the function getDissimilarity(). The
divergence
measure is sensitive to sample size. Subsampling or bootstrapping can be
applied to equalize sample sizes between comparisons.
data(GlobalPatterns)
tse <- GlobalPatterns
# By default, reference is median of all samples. The name of column where
# results is "divergence" by default, but it can be specified.
tse <- addDivergence(tse)
# The method that are used to calculate distance in divergence and
# reference can be specified. Here, euclidean distance is used. Reference is
# the first sample. It is recommended # to add reference to colData.
tse[["reference"]] <- rep(colnames(tse)[[1]], ncol(tse))
tse <- addDivergence(
tse, name = "divergence_first_sample",
reference = "reference",
method = "euclidean")
# Here we compare samples to global mean
tse <- addDivergence(tse, name = "divergence_average", reference = "mean")
# All three divergence results are stored in colData.
colData(tse)
#> DataFrame with 26 rows and 11 columns
#> X.SampleID Primer Final_Barcode Barcode_truncated_plus_T
#> <factor> <factor> <factor> <factor>
#> CL3 CL3 ILBC_01 AACGCA TGCGTT
#> CC1 CC1 ILBC_02 AACTCG CGAGTT
#> SV1 SV1 ILBC_03 AACTGT ACAGTT
#> M31Fcsw M31Fcsw ILBC_04 AAGAGA TCTCTT
#> M11Fcsw M11Fcsw ILBC_05 AAGCTG CAGCTT
#> ... ... ... ... ...
#> TS28 TS28 ILBC_25 ACCAGA TCTGGT
#> TS29 TS29 ILBC_26 ACCAGC GCTGGT
#> Even1 Even1 ILBC_27 ACCGCA TGCGGT
#> Even2 Even2 ILBC_28 ACCTCG CGAGGT
#> Even3 Even3 ILBC_29 ACCTGT ACAGGT
#> Barcode_full_length SampleType
#> <factor> <factor>
#> CL3 CTAGCGTGCGT Soil
#> CC1 CATCGACGAGT Soil
#> SV1 GTACGCACAGT Soil
#> M31Fcsw TCGACATCTCT Feces
#> M11Fcsw CGACTGCAGCT Feces
#> ... ... ...
#> TS28 GCATCGTCTGG Feces
#> TS29 CTAGTCGCTGG Feces
#> Even1 TGACTCTGCGG Mock
#> Even2 TCTGATCGAGG Mock
#> Even3 AGAGAGACAGG Mock
#> Description divergence reference
#> <factor> <numeric> <character>
#> CL3 Calhoun South Carolina Pine soil, pH 4.9 0.989114 CL3
#> CC1 Cedar Creek Minnesota, grassland, pH 6.1 0.991217 CL3
#> SV1 Sevilleta new Mexico, desert scrub, pH 8.3 0.986994 CL3
#> M31Fcsw M3, Day 1, fecal swab, whole body study 0.995435 CL3
#> M11Fcsw M1, Day 1, fecal swab, whole body study 0.996395 CL3
#> ... ... ... ...
#> TS28 Twin #1 0.991388 CL3
#> TS29 Twin #2 0.992698 CL3
#> Even1 Even1 0.990063 CL3
#> Even2 Even2 0.989827 CL3
#> Even3 Even3 0.991461 CL3
#> divergence_first_sample divergence_average
#> <numeric> <numeric>
#> CL3 0.0 0.879196
#> CC1 83210.0 0.875744
#> SV1 73809.5 0.915286
#> M31Fcsw 419594.0 0.842727
#> M11Fcsw 626574.7 0.870541
#> ... ... ...
#> TS28 185596 0.813599
#> TS29 352153 0.863493
#> Even1 225268 0.809229
#> Even2 194434 0.808371
#> Even3 204304 0.814546