Normalize OTU Table by Copy Number

One way to correct OTU abundance tables is to take the rRNA copy number of the detected species into account and divide the detected read number for each OTU by the rRNA copy number. This can be done with the Normalize OTU Table by Copy Number (beta) tool. Note that this algorithm corrects the species distribution for each sample individually to get a more realistic picture of the actual species distribution in a sample. In order to normalize OTU abundance tables across samples one can use the Create Normalized Abundance Subtable button (see Visualization of OTU abundance tables), but many tools use an internal cross-sample normalization strategy.

In order to run this tool, an Amplicon Multiplication Table is required. Such tables can be imported with Import PICRUSt2 Multiplication Table (beta).

To run the tool, go to

        Toolbox | Microbial Genomics Module (Image mgm_folder_closed_flat_16_h_p) | Metagenomics (Image wma_folder_open_flat_16_n_p) | Amplicon-Based Analysis (Image otutools_open_16_n_p) | Normalize OTU Table by Copy Number (beta) (Image normalized_otu_16_n_p)

Select the OTU table you want to normalize first and click "Next".

In the next wizard step (see figure 5.1) you can choose the Amplicon Multiplication Table to use. Image normalize_otu_table_wizard1
Figure 5.1: Selecting an Amplicon Multiplication Table for normalizing OTU tables.

The normalization works in the same way as the functional inference, except for that the functional inference step is left out, see Inference of functional abundances from 16S/ITS data.