PseudotimeMSProcess

Perform multi-sample pattern integration for Pseudotimecascade results

Usage

PseudotimeMSProcess(integrate_list, conf = 0.75, qval = 0.05)

Arguments

integrate_list

A named list of single-sample Pseudotimecascade results. Each element must contain gene_group and fit_data_list.

conf

Minimum confidence threshold for consensus pattern selection (default = 0.75).

qval

q-value cutoff for filtering significant genes (default = 0.05).

Value

A list (hereafter referred to as gene_mean_list) containing:

mean_expr

Matrix of averaged fitted expression across samples.

mean_pattern

Gene pattern result based on mean trajectory.

df_pattern

Consensus pattern table with confidence scores.

df_qvalue

Filtered q-value matrix aligned to consensus genes.

df_switch_point

List of switch point interval matrices across samples.

Details

This function integrates single-sample Pseudotimecascade results stored in a list and performs consensus pattern analysis across samples. For each gene, the most frequent expression pattern is identified, and a confidence score is computed as the proportion of samples supporting the dominant pattern. Genes with confidence greater than or equal to conf and passing the q-value threshold are retained.

For retained genes, the function:

• Computes the average fitted expression trajectory across valid samples.

• Reassigns gene patterns based on the mean trajectory.

• Collects sample-level switch points and organizes them into interval matrices.

Author

Zhicheng Ji, Changxin Wan, Beijie Ji