31. dmc_ttest

31.1. Overview

dmc_ttest performs differential CpG analysis on DNA methylation Beta-values using parametric tests.

The test is selected from the number of groups and the requested two-group test mode:

Comparison	Test
Two groups	Student’s two-sample t-test by default.
Two groups with `--welch`	Welch’s two-sample t-test.
Two groups with `--paired`	Paired t-test.
Three or more groups	One-way ANOVA.

The input matrix must have CpGs/probes in rows and samples in columns.

31.2. Input Files

31.2.1. Beta-value matrix

The first row contains sample IDs and the first column contains CpG/probe IDs.

Example:

CpG_ID      Sample_01   Sample_02   Sample_03   Sample_04
cg00001099  0.775       0.812       0.623       0.598
cg00000363  0.611       0.602       0.470       0.455

Sample IDs must be unique.

Non-numeric and non-finite values are treated as missing values. Plain-text, .gz, .bz2, and other formats supported by the CpGtools reader may be used.

31.2.2. Group file

The group file is a comma-separated, two-column file with a header:

column 1: sample ID
column 2: group ID

At least two groups are required.

Example:

Sample,Group
Sample_01,Control
Sample_02,Control
Sample_03,Case
Sample_04,Case

Sample IDs in the group file must be unique.

Samples listed in the group file but absent from the methylation matrix are reported and excluded from the analysis. Samples present in the matrix but not listed in the group file are ignored.

31.3. Two-group Tests

31.3.1. Student’s t-test

With two groups and no additional test flag, dmc_ttest uses the standard independent two-sample t-test with equal variances assumed.

For a CpG to be tested, each group must contain at least two valid observations.

31.3.2. Welch’s t-test

Use --welch for an independent two-sample t-test that does not assume equal variances:

dmc_ttest \
    -i test_05_TwoGroup.tsv.gz \
    -g test_05_TwoGroup.grp.csv \
    --welch \
    -o ttest_welch

31.3.3. Paired t-test

Use --paired when the two groups contain matched samples:

dmc_ttest \
    -i paired_beta.tsv \
    -g paired_groups.csv \
    --paired \
    -o ttest_paired

Important behavior:

exactly two groups are required;
the groups must contain the same number of samples after matching to the methylation matrix;
samples are paired by their order within each group;
missing values are removed pairwise for each CpG; and
at least two complete pairs are required for a valid test.

If both --paired and --welch are supplied, --welch is ignored.

31.4. Multi-group ANOVA

When three or more groups are defined, dmc_ttest automatically performs a one-way ANOVA using scipy.stats.f_oneway.

Each group must contain at least one valid observation for a CpG. If any group has no valid value for that CpG, the ANOVA result is reported as missing.

31.5. Delta Beta

For two-group comparisons, delta_beta is calculated as:

\[\Delta\beta = \mathrm{mean}(\mathrm{group}_1) - \mathrm{mean}(\mathrm{group}_2)\]

The group IDs are sorted internally, so group 1 is the first group in sorted order and group 2 is the second.

For three or more groups, delta_beta is not defined and is written as n/a.

31.6. Usage

Standard two-group t-test:

dmc_ttest \
    -i test_05_TwoGroup.tsv.gz \
    -g test_05_TwoGroup.grp.csv \
    -o ttest_2G

Three-group ANOVA:

dmc_ttest \
    -i test_06_ThreeGroup.tsv.gz \
    -g test_06_ThreeGroup.grp.csv \
    -o ttest_3G

Useful options include:

-p, --paired – use a paired t-test for exactly two groups
-w, --welch – use Welch’s t-test for exactly two independent groups
-i, --input_file – Beta-value matrix
-g, --group – sample/group file
-o, --output – output prefix
--version – show the CpGtools version

Display all options with:

dmc_ttest -h

31.7. Output

For output prefix ttest_2G, the command writes:

ttest_2G.pval.txt

The original input table is preserved and three columns are appended:

Column	Description
`delta_beta`	Difference between the two group means. Written as `n/a` for multi-group ANOVA or when it cannot be calculated.
`pval`	Raw p-value from the selected t-test or one-way ANOVA.
`adj.pval`	Benjamini-Hochberg FDR-adjusted p-value.

Missing p-values and adjusted p-values are written as n/a and excluded from multiple-testing correction.

31.8. Input-row Handling

If a CpG row contains fewer data fields than the header, missing values are padded with NaN. Extra fields are ignored. A warning is reported in either case.

The original input row is preserved in the final output; only delta_beta, pval, and adj.pval are appended.

31.9. Example Data

Two groups:

Three groups: