Zeta: Calculation of zeta and weighted zeta score.

In ZetaSuite: Analyze High-Dimensional High-Throughput Dataset and Quality Control Single-Cell RNA-Seq

Calculation of zeta and weighted zeta score.

Description

This function calculates zeta scores for genes based on their Z-score profiles across different thresholds. The zeta score quantifies the regulatory effect of gene knockdown on alternative splicing events by measuring the area under the curve of event coverage across Z-score thresholds.

Usage

Zeta(ZscoreVal, ZseqList, SVMcurve = NULL, SVM = FALSE)

Arguments

ZscoreVal	A matrix of Z-scores where rows represent genes and columns represent readouts/conditions. This is typically the output from the Zscore() function.
ZseqList	A list containing two vectors: 'Zseq_D' (decrease direction thresholds) and 'Zseq_I' (increase direction thresholds). These define the Z-score bins for calculating event coverage.
SVMcurve	Optional. A matrix containing SVM curve data for decrease and increase directions. Required only when SVM=TRUE. The matrix should have 4 columns: Z-score and coverage for decrease direction (columns 1-2), and Z-score and coverage for increase direction (columns 3-4).
SVM	Logical. Whether to use SVM curves for background correction. Default is FALSE. When TRUE, the function subtracts SVM-predicted background from the event coverage before calculating zeta scores.

Details

The function calculates zeta scores as follows:

1. For each Z-score threshold, calculates the proportion of readouts that exceed (increase) or fall below (decrease) the threshold

2. Computes the area under the event coverage curve using trapezoidal integration

3. If SVM=TRUE, subtracts SVM-predicted background coverage before area calculation

4. Returns separate scores for decrease (Zeta_D) and increase (Zeta_I) directions

Higher zeta scores indicate stronger regulatory effects on alternative splicing.

Value

A data frame with two columns:

Zeta_D	Zeta score for decrease direction (exon skipping events)
Zeta_I	Zeta score for increase direction (exon inclusion events)

Each row corresponds to a gene, and the zeta scores represent the cumulative regulatory effect across all Z-score thresholds.

Author(s)

Yajing Hao, Shuyang Zhang, Junhui Li, Guofeng Zhao, Xiang-Dong Fu

Examples

data(ZseqList)
data(SVMcurve)
data(countMat)
data(negGene)
ZscoreVal <- Zscore(countMat, negGene)
zetaData <- Zeta(ZscoreVal, ZseqList, SVM=FALSE)

ZetaSuite documentation built on Nov. 5, 2025, 6:37 p.m.