Shibiao Wan

MOTLAB improves breast cancer outcome prediction in underrepresented Black women by integrating multi-omics data with transfer learning and weighting strategies.

DAG-HEART improves breast cancer outcome prediction in underrepresented populations by integrating multi-omics data with directed acyclic graph constraints and transfer learning.

The paper develops a meta-clustering framework to identify high-resolution molecular subtypes of pediatric low-grade glioma, revealing significant transcriptomic heterogeneity across five subgroups.

SAMP V2 improves antimicrobial peptide identification by integrating biochemical and sequence-order information into split amino acid composition, outperforming existing models on multiple datasets.

MetaAMI uses a meta-learning framework with random projections to predict in-hospital mortality in AMI patients, outperforming baseline classifiers across multiple metrics.

PRIME integrates single-cell and spatial transcriptomics data at the atlas level, improving batch correction and preserving biological variation across diverse datasets.

OTTER improves T-ALL molecular characterization by integrating transcriptomic and genomic data using optimal transport, outperforming existing methods and identifying subtype-specific features.

PalmaClust detects ultra-rare cell types in scRNA-seq by fusing graphs based on the Palma ratio, improving rare-class F1 scores by 20% over existing methods.

RaMBat accurately identifies medulloblastoma subtypes from diverse data sources with severe batch effects, achieving 99% median accuracy.

MetaPaCS improves pancreatic cancer subtype identification using a meta-learning framework that combines multiple base classifiers, outperforming individual models and existing methods.

The paper proposes a multi-omics transfer learning framework to reduce racial disparities in prostate adenocarcinoma prognosis by integrating mRNA, miRNA, and methylation data.

This study shows ECD and ERBB2 cooperate in breast cancer by enhancing unfolded protein response and glycolysis through direct mRNA binding and increased glucose uptake.

AttentionAML is a deep learning framework for AML subtype classification using transcriptomic data, achieving high accuracy and outperforming existing methods.

RPSLearner improves NSCLC subtype classification using random projection and stacking learning, achieving higher accuracy and AUC than existing methods.

The paper uses machine learning to analyze functional connectivity in brain networks to distinguish cocaine use disorder from healthy controls, finding that SN and CEN are more critical than DMN in addiction.

SAMP identifies antimicrobial peptides using an ensemble model with new features, outperforming existing methods on multiple datasets.

WIMOAD integrates multi-omics data using a meta learning framework to improve Alzheimer's disease diagnosis by capturing complex patterns linked to clinical outcomes.

RanBALL is a model for B-ALL subtype identification using random projection and ensemble learning, achieving high accuracy and outperforming existing methods.

This study shows YAP/TAZ are essential for maintaining radial glial cell function during brain development, affecting neuronal and ependymal cell numbers and causing hydrocephaly when disrupted.

SHARP processes large-scale scRNA-seq data efficiently, outperforming existing methods in speed and accuracy for datasets over 40,000 cells.