Katerina J. Kechris

MT-LLE: Multi-Task Locally Linear Embedding for Interpretable Disease Modeling from Longitudinal Omics Data

This paper introduces MT-LLE, a multi-task manifold learning method that improves disease modeling from longitudinal omics data by integrating clinical labels and temporal progression, achieving better clinical structure and progression forecasting than standard methods.

MetaboXcan: A multiomic framework linking genetically predicted metabolites, gene expression, and complex traits

MetaboXcan predicts plasma metabolites from genetic data and associates them with complex traits, identifying disease risk genes and metabolites in chronic kidney disease.

Differences in immune cell profiles around the time of islet autoimmunity seroconversion in children with and without type 1 diabetes

The paper examines immune cell profiles around islet autoimmunity seroconversion in children with and without type 1 diabetes, finding differences in immune cell ratios that may reflect immune maturation disruptions.

A Generalized Higher-order Correlation Analysis Framework for Multi-Omics Network Inference

This paper introduces SGTCCA-Net, a framework for multi-omics network inference that addresses limitations of existing methods by incorporating higher-order correlations and improving computational efficiency.

Matrix Linear Models for connecting metabolite composition to individual characteristics

The paper develops a matrix linear model to connect metabolite composition with individual characteristics, showing it can separate overlapping metabolite features.

Prediction and Characterization of Genetically Regulated Expression of Target Tissues in Asthma

The paper develops gene expression prediction models for asthma-related tissues in African-ancestry populations and identifies 17 candidate causal asthma genes using transcriptome-wide association study.

Proteomic Networks and Related Genetic Variants Associated with Smoking and Chronic Obstructive Pulmonary Disease

The paper identifies proteomic networks and genetic variants associated with smoking and COPD phenotypes using multi-ethnic proteomic data and network-based GWAS.

SmCCNet 2.0: A Comprehensive Tool for Multi-omics Network Inference

SmCCNet 2.0 integrates multi-omics data with a phenotype to reconstruct networks specific to that phenotype, offering improved setup flexibility and user experience.

PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration

PathIntegrate integrates multi-omics data using pathway-based modeling to identify important pathways and molecules contributing to outcomes, demonstrated on COPD and COVID-19 datasets.

RCFGL: Rapid Condition adaptive Fused Graphical Lasso and application to modeling brain region co-expression networks

This paper introduces RCFGL, a fast method for joint estimation of condition-specific gene co-expression networks, showing improved computational efficiency and performance over existing methods.

Proteomic Fingerprinting: A novel privacy concern

This study shows that proteomic data can be linked to specific genomes using pQTL knowledge, highlighting privacy risks and the need for protections similar to genomic data.

PaIRKAT: A pathway integrated regression-based kernel association test with applications to metabolomics and COPD phenotypes

PaIRKAT improves testing power for pathway-phenotype associations by integrating graph structure into kernel regression, showing effectiveness in metabolomics data.

Analysis of SARS-CoV-2 Mutations Over Time Reveals Increasing Prevalence of Variants in the Spike Protein and RNA-Dependent RNA Polymerase

Condition-adaptive fused graphical lasso (CFGL): an adaptive procedure for inferring condition-specific gene co-expression network

The paper introduces CFGL, a method for inferring condition-specific gene co-expression networks, showing improved accuracy and revealing biological insights in rat and cancer datasets.