Description Usage Arguments Value

Permutation-based differential abundance analysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 | ```
ZicoSeq(
meta.dat,
comm,
grp.name,
adj.name = NULL,
prev.filter = 0.1,
abund.filter = 10,
min.prop = 0,
is.winsor = TRUE,
winsor.qt = 0.97,
is.prior = TRUE,
prior.dist = c("BetaMix", "ZIBB"),
post.method = c("sample", "mean"),
post.sample.no = 25,
link.func = list(function(x) x^0.25, function(x) x^0.5, function(x) x^0.75),
link.d.func = list(function(x) 0.25 * x^(-0.75), function(x) 0.5 * x^(-0.5),
function(x) 0.75 * x^(-0.25)),
variance.EB = FALSE,
df.prior = 10,
perm.no = 99,
strata = NULL,
stats.combine.func = max,
stage.no = 6,
topK = NULL,
stage.fdr = 0.75,
stage.max.pct = 0.5,
is.fwer = FALSE,
is.tree.fdr = FALSE,
tree = NULL,
verbose = TRUE,
return.comm = FALSE,
return.perm.F = FALSE,
...
)
``` |

`meta.dat` |
a data frame containing the sample information |

`comm` |
a matrix of counts, row - features (OTUs, genes, etc) , column - sample |

`grp.name` |
a character, variable of interest; it could be numeric or categorical; should be in "meta.dat" |

`adj.name` |
a character vector, variable(s) to be adjusted; they could be numeric or categorical; should be in "meta.dat" |

`prev.filter` |
features with prevalence (i.e., nonzero proportion) less than "prev.cutoff" or be filtered |

`abund.filter` |
features with a total counts less than "abund.cutoff" or be filtered |

`min.prop` |
Undetermined |

`is.winsor` |
a logical value indicating whether winsorization should be performed to replace outliers. The default is TRUE. |

`winsor.qt` |
the winsorization quantile, above which the counts will be replaced |

`is.prior` |
a logical value indicating whether to perform posterior inference based on some prior distribution on the proportion data |

`prior.dist` |
prior distribution, either two-component beta-binomial mixture ("BetaMix") or zeroinflated beta-binomial ("ZIBB") |

`post.method` |
method for posterior inference, either based on posterior sampling ("sample") or approximate posterior mean ("mean") |

`post.sample.no` |
the number of posterior samples if posterior sampling is used |

`link.func` |
a list of functions that connects the ratios to the covariates |

`link.d.func` |
a list of the derivative function of "link.func"; only need to specifiy when "post.method" is "mean" |

`variance.EB` |
a logical value indicating whehter to perform empirical Bayes based variance shrinkage |

`df.prior` |
the degree of freedom of the prior inverse gamma distribution for variance shrinkage |

`perm.no` |
the number of permutations; If the raw p values are of the major interest, set "perm.no" to at least 999 |

`strata` |
a factor indicating the permutation strata; permutation will be confined to each stratum |

`stats.combine.func` |
function to combine the F-statistic for the omnibus test |

`stage.no` |
the number of stages if multiple-stage ratio stategy is used |

`topK` |
the number of dominant features that will be excluded in the initial stage ratio calculation |

`stage.fdr` |
the fdr cutoff below which the features will be excluded for calculating the ratio |

`stage.max.pct` |
the maximum percentage of features that will be excluded |

`is.fwer` |
a logical value indicating whether the family-wise error rate control (West-Young) should be performed |

`is.tree.fdr` |
a logical value indicating whether tree-based false discovery rate shuold be carried out |

`tree` |
a class of "phylo", the tree relats all the OTUs, and should have the same names in "comm" |

`verbose` |
a logical value indicating whether the trace information should be printed out |

`return.comm` |
a logical value indicating whether the wisorized, filtered "comm" matrix should be returned |

`return.perm.F` |
Undetermined |

`...` |
arguments passing to tree-based fdr control |

A list with the elements

`call` |
the call |

`comm` |
the wisorized, filtered "comm" matrix |

`filter.ind` |
a vector of logical values indicating which features are tested |

`R2` |
a matrix of percent explained variance (number of features by number of functions) |

`F0` |
a matrix of F-statistics (number of features by number of functions) |

`RSS` |
a matrix of residual sum squares (number of features by number of functions) |

`df.model, df.residual` |
degree of freedoms for the model and residual space |

`p.raw` |
the raw p-values based on permutations (not accurate if "perm.no" is small) |

`p.adj.fdr` |
permutation-based FDR-adjusted p-values |

`p.adj.tree.fdr` |
permutation-based tree FDR-adjusted p-values |

`p.adj.fwer` |
permutation-based FWER-adjusted (West-Young) p-values |

`tree.fdr.obj` |
the object returned by the "TreeFDR" |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.