scSniper: Single-cell Deep Neural Network-based Identification of Prominent Biomarkers
Li, M.; Chen, Y.; Ding, J.
Show abstract
Discovering disease biomarkers at the single-cell level is crucial for advancing our understanding of diseases and improving diagnostic accuracy. However, current computational methods often have limitations, such as a reliance on prior knowledge, constraints to unimodal data, and the use of conventional statistical tests for feature selection. To address these issues, we introduce scSniper, a novel approach that employs a specialized deep neural network framework tailored for robust single-cell multiomic biomarker detection. A standout feature of scSniper is the mimetic attention block, enhancing alignment across multi-modal data types. Moreover, scSniper utilizes sensitivity analysis based on a deep neural network for feature selection and uncovers intricate gene regulatory networks without requiring prior knowledge. Comprehensive evaluations on real-world datasets, including COVID-19 CITE-Seq and LUAD scRNA-Seq, demonstrate scSnipers exceptional ability to identify critical biomarkers consistently outperforming traditional methods like MAST, Wilcox, and DESeq2. The scSniper tool and related experimental codes are publicly accessible at https://github.com/mcgilldinglab/scSniper.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- StereoMM: A Graph Fusion Model for Integrating Spatial Transcriptomic Data and Pathological Images 96%
- SHEST: Single-cell-level artificial intelligence from haematoxylin and eosin morphology for cell type prediction and spatial transcriptomics reconstruction 95%
- scDeepInsight: a supervised cell-type identification method for scRNA-seq data with deep learning 95%
Similar papers in this journal
- CellFM: a large-scale foundation model pre-trained on transcriptomics of 100 million human cells 96%
- scGCN: a Graph Convolutional Networks Algorithm for Knowledge Transfer in Single Cell Omics 96%
- MORONET: Multi-omics Integration via Graph Convolutional Networks for Biomedical Data Classification 96%
Similar papers in this journal
- Bi-level Graph Learning Unveils Prognosis-Relevant Tumor Microenvironment Patterns in Breast Multiplexed Digital Pathology 96%
- scTenifoldNet: a machine learning workflow for constructing and comparing transcriptome-wide gene regulatory networks from single-cell data 95%
- Obtaining Spatially Resolved Tumor Purity Maps Using Deep Multiple Instance Learning In A Pan-cancer Study 95%
Similar papers in this journal
Similar papers in this journal
- GREMI: an Explainable Multi-omics Integration Framework for Enhanced Disease Prediction and Module Identification 96%
- MOH: a novel multilayer multi-omics heterogeneous graph for single-cell clustering 96%
- pathCLIP: Detection of Genes and Gene Relations from Biological Pathway Figures through Image-Text Contrastive Learning 95%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.