Testing 4 levels against a control but the results show only two levels

Downstream analysis and statistics MaAsLin
1

I’m testing the effect of 4 dietary groups against a control (initial state).
Strangely, the summary plot I get from Maaslin3 shows beta-coef only for two of the levels (0 and 50) and prevalence and abundance only for the two other levels (25 and 75).

Why don’t I get the analysis results for all levels?

My call:

maaslin3(input_data = otu_table(Ps_obj),
                    input_metadata = get_variable(Ps_obj),
                    output = 'MaASLin3_output',
                    formula = paste("~", Dietary.group),
                    normalization = 'TSS',
                    transform = 'LOG',
                    augment = TRUE,
                    standardize = TRUE,
                    max_significance = 0.1,
                    median_comparison_abundance = TRUE,
                    median_comparison_prevalence = FALSE,
                    max_pngs = 250,
                    cores = 1)

The output:

summary_plot
summary_plot3324×3330 409 KB

Thanks in advance,

Roey

2

Hi @rangel,

It looks like the plot you posted does show all levels (0, 50, 25, 75). The 25 and 75 groups are shown in the heatmap portion of the plot.

MaAsLin’s summary plot will automatically choose the groups with the most significant effects to be shown in the two dot plots whereas the others will be shown in the heatmap. You can change how this is shown following the instructions in the user manual.

I would also recommend checking your results.tsv as it should include all the levels you were interested in.

Thanks,
Jacob Nearing

1 Like
3

Thanks! I missed the coef_plot_vars and heatmap_vars options