Single cell pathway analysis method based on Overlapping Group Lasso development

By using the Logistics form and Stability Selection algorithm of the Overlapping Group Lasso model in pathway analysis, the existing methods ignore overlapping gene effects and low computational efficiency are solved, and more accurate and efficient pathway screening is achieved, improving the interpretability and predictive performance of single-cell data analysis.

CN119694408AActive Publication Date: 2025-03-25ACAD OF MATHEMATICS & SYSTEMS SCIENCE - CHINESE ACAD OF SCI
5 Cites 0 Cited by

Patent Information

Application Number
CN202411634758.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-15
Publication Date
2025-03-25
Estimated Expiration
2044-11-15

AI Technical Summary

Technical Problem

The existing pathway analysis methods ignore the overlapping effect of the same genes between different genomes, resulting in inaccurate screening results, and the Overlapping Group Lasso model is slow to calculate under large data volumes.

Method used

The Logistics form of the Overlapping Group Lasso model combined with the Stability Selection algorithm was used to screen the gene expression data of the two cell types by pathway, decomposing the effects of overlapping genes, and improving the stability of the screening results through multiple sampling and parameter tuning.

Benefits of technology

It effectively avoids the influence of overlapping gene effects, improves the accuracy of pathway screening, reduces the subjective control of model hyperparameters, and obtains more than 90% prediction accuracy on simulated and real data sets.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a single cell pathway analysis method based on Overlapping Group Lasso development. The method comprises the following steps: carrying out regression on gene expression data and a cell state of each cell by using a Logistic form of a model; a Stability Selection algorithm is adopted, multiple times of sampling regression are carried out on the data, and a path with the most stable screening frequency is reserved to serve as a final result; the method specifically comprises the following steps: 1, carrying out mathematical expression and preprocessing on an input cell gene expression matrix, pathway data and a cell label; 2, establishing a regression model; 3, solving a regression model; and 4, stability selection. According to the method, an independent variable group taking a channel as a unit is screened out by the model, and based on the characteristics of Overlapping Group Lasso, the effect of overlapped genes in different genomes is decomposed, so that the influence caused by the effect of the overlapped genes is effectively avoided.
Need to check novelty before this filing date? Find Prior Art