delta_BD {HTSSIP}

delta_BD calculation

Description

Calculate delta_BD as described in Pepe-Ranney et al., 2016.

Usage

delta_BD(physeq, control_expr, n = 20, BD_min = NULL, BD_max = NULL)

Arguments

physeq

Phyloseq object

control_expr

An expression for identifying unlabeled control samples in the phyloseq object (eg., "Substrate=='12C-Con'")

n

How many evenly-spaced buoyant density values to use for linear interpolation of abundances.

BD_min

The minimum BD value of the BD range used for OTU abundance interpolation. If NULL, then BD_min will be the minimum of all BD values in the phyloseq object.

BD_max

The maximum BD value of the BD range used for OTU abundance interpolation. If NULL, then BD_max will be the maximum of all BD values in the phyloseq object.

Details

Basically, the abundance of each OTU is interpolated at specific BD values in order to have abundance values at consistent points across gradients (gradient fraction BDs normally vary from gradient to gradient). The center of mass (CM) is calculcated from these interpolated values, which is the weighted mean BD with interpolated OTU abundances used as weights (ie., where in the density gradient contains the 'center' of the OTU abundance distribution). Delta_BD is then calculated by substracting the CM for the unlabeled control gradient from the labeled treatment gradient.

The delta_BD calculation will be a comparison between unlabled control and labeled treatment samples. These samples are distinguished from each other with the 'control_expr' parameter. NOTE: if multiple gradients fall into the control or treatment category, they will be treated as one gradient (which may be OK if you want to combine replicate gradients).

NaN values may occur due low abundances.

The BD range used for interpolation is set by the min/max of all buoyant density values in the phyloseq object (standardize across).

Value

data.frame with delta_BD values for each OTU. 'CM' stands for 'center of mass'.

Examples

data(physeq_S2D2_l)
# just selecting 1 treatment-control comparison
physeq = physeq_S2D2_l[[1]]

## Not run: 
# calculating delta_BD
df = delta_BD(physeq, control_expr='Substrate=="12C-Con"')
head(df)

# In this example, the replicate gradients will be combined for treatments/controls
data(physeq_rep3)
df = delta_BD(physeq_rep3, control_expr='Treatment=="12C-Con"')
head(df)

## End(Not run)