Statistical Comparisons

DPKS allows for a number of different statistical tests to be performed between experimental groups in your data. Currently, it is possible to compare your samples using:

1. T-test - 2-sided t-test for the independent samples.
2. Linear Regression - 2-sided test to calculate a linear least-squares regression for the abundances betweeen experimental groups.
3. ANOVA - One-way ANOVA to compare the means of 2 groups.
4. Paired T-test - 2-sided t-test for 2 related samples (need to indicate a Pairs column in the Design Matrix).

T-test

qm = qm.compare(
    method="ttest",
    comparisons=(2,1),
    min_samples_per_group=10,
    level="protein",
    multiple_testing_correction_method="fdr_tsbh"
)

Linear Regression

qm = qm.compare(
    method="linregress",
    comparisons=(2,1),
    min_samples_per_group=10,
    level="protein",
    multiple_testing_correction_method="fdr_tsbh"
)

ANOVA

qm = qm.compare(
    method="anova",
    comparisons=(2,1),
    min_samples_per_group=10,
    level="protein",
    multiple_testing_correction_method="fdr_tsbh"
)

Paired T-test

In order to perform paired t-tests with your data, you first need to pass in a "Pair" column with your design matrix:

Sample	Group	Pair
s1	1	s2
s2	2	s1
s3	2	s4
s4	1	s3

Here, you need to make sure that your Sample is paired with another valid Sample in the list, but each sample pairing should be unique.

qm = qm.compare(
    method="ttest_paired",
    comparisons=(2,1),
    min_samples_per_group=10,
    level="protein",
    multiple_testing_correction_method="fdr_tsbh"
)

Multiple Comparisons

It is possible to perform multiple comparisons if you have multiple groups in your data by passing a list of tuples in as the comparison parameter:

qm = qm.compare(
    method="ttest_paired",
    comparisons=[(2,1), (3, 1), (4, 1)],
    min_samples_per_group=10,
    level="protein",
    multiple_testing_correction_method="fdr_tsbh"
)

The above will separately compare groups 2, 3, and 4 to group 1 and write results columns for each of the 3 different comparisons.

Example

There is a jupyter notebook with some examples of how to use this functionality and some possible plots.

Differential Expression: Demonstrates how to compute differences between two experimental conditions.