Grace-AKO: a novel and stable knockoff filter for variable selection incorporating gene network structures.
Peixin Tian1, Yiqian Hu1, Zhonghua Liu2
1Department of Statistics and Actuarial Science, The University of Hong Kong, Hong Kong SAR, China.
This study introduces Grace-AKO, a new method for selecting important genes in high-dimensional data. It effectively controls false discovery rates (FDR) and identifies key genes, improving upon existing models.
Area of Science:
- Genomics
- Biostatistics
- Bioinformatics
Background:
- Variable selection is crucial for identifying genes linked to clinical outcomes in genomic studies.
- High-dimensional data with limited samples necessitates methods to control false discovery rate (FDR).
Purpose of the Study:
- To propose a novel method, Grace-AKO, for graph-constrained estimation to identify significant genes.
- To control finite-sample FDR and enhance model stability in high-dimensional genomic data.
Main Methods:
- Grace-AKO combines aggregation of multiple knockoffs (AKO) with a network-constrained penalty.
- The method is designed to control FDR in finite-sample settings.
Main Results:
- Simulation studies demonstrate Grace-AKO's superior finite-sample FDR control compared to the original Grace model.
- Application to prostate cancer data identified 47 candidate genes associated with prostate-specific antigen (PSA) levels.
- Over 75% of the identified genes were validated, indicating strong biological relevance.
Conclusions:
- Grace-AKO offers an effective approach for gene selection in high-dimensional data.
- The method successfully controls FDR and improves model stability, with validated findings in a real-world cancer dataset.
More Related Videos
07:11Author Spotlight: Emerging Technologies and Advanced Tools for Decoding Metabolomics Data Analysis
Published on: November 10, 2023
09:33Author Spotlight: Finding New Therapeutic Targets for Malignant Peripheral Nerve Sheath Tumor Through Genome-Scale shRNA Screens
Published on: August 25, 2023
Related Concept Videos
Combinatorial Gene Control
The expression of more than 30,000 genes is controlled by approximately 2000-3000 transcription factors. This is possible because a single transcription factor can recognize more than one regulatory sequence. The specificity in gene...
Genetic Screens
Forward genetic screens
Forward or “classical” genetic screens involve creating random mutations in an organism’s DNA using radiation, mutagens, or insertion of additional bases, which...
Frequency-dependent Selection
Genetic Variation
Genes exist in different versions called alleles,...
Protein Networks
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
One-Compartment Open Model: Wagner-Nelson and Loo Riegelman Method for ka Estimation
On...
