Data Combination

In this section, we will cover the key function in MicrobiomeStat for combining multiple microbiome datasets into one object for integrated analysis.

Data Combination

Combining multiple microbiome datasets is a crucial step in many research projects. The mStat_combine_data() function in MicrobiomeStat provides a straightforward way to merge two datasets into one object for integrated analysis.

Overview

mStat_combine_data() is designed specifically for combining two MicrobiomeStat data objects that are in the raw format.

Each data object should contain:

• feature.tab: OTU/ASV table

• meta.dat: Sample metadata

• feature.ann: Taxonomic annotations

The function will:

• Row-bind the two feature.tab matrices

• Row-bind the two feature.ann matrices

• Row-bind the two meta.dat data frames

The output is a single merged data object ready for integrated analysis.

Usage

mStat_combine_data(
  data.obj1 = obj1, 
  data.obj2 = obj2
)

• data.obj1: The first data object to combine

• data.obj2: The second data object to combine

Details

Here is how mStat_combine_data() works under the hood:

First it checks that both input objects are in the raw MicrobiomeStat format.

It then identifies common features and samples between the two objects.

• For common features, it checks that the data values are consistent between the two objects.

• If no common features or samples are found, it will print a warning message.

Next, it performs:

• A full join of the two feature.tab matrices by feature IDs

• Replaces any NA values with 0

• Gathers into a long format

• Spreads back to wide format

• Sets feature IDs as row names

This yields the combined feature.tab matrix.

It performs similar operations to combine feature.ann and meta.dat.

Finally, it returns the merged data object containing all three components.

Applications

Key applications of this function include:

• Merge case and control groups for differential abundance analysis

• Combine multiple cohorts for meta-analysis

• Compile temporal datasets for longitudinal analysis

• Create one large dataset from multiple studies of a population

Proper dataset integration is crucial for maximizing the potential of your microbiome research. mStat_combine_data() lets you seamlessly merge compatible data objects, enabling more powerful integrated analysis.