Kathryn Roeder

This study identifies three functional programs impacted by autism genes, contextualized across neuronal subclasses, developmental timing, and cortical regions.

PAQu estimates protein isoform abundances using multiomic data, outperforming existing methods and revealing increased C4A isoform levels in schizophrenia.

The paper identifies 253 autism-associated genes using rare variation from 62,429 individuals and finds pleiotropy across physiological systems and developmental disorders.

The paper introduces TIDEST, a framework that improves statistical inference in reconstructed spatial transcriptomes by correcting for prediction errors and spatial confounding, leading to more accurate biological interpretations.

LIKA, a likelihood model over many-to-many kinase-substrate networks, beats KSEA in simulations with shared substrates, recovers EGFR and ABL1 in cell lines, and ranks MAPK3 first in schizophrenia cortex.

The paper uses STARMAPS to analyze single-cell RNA sequencing data and identifies spatial and temporal patterns of ASD-associated gene enrichment in specific brain regions.

The paper develops a model to analyze X-linked rare variation in ASD, identifying 9 genes associated with ASD and explaining how X-linked variation could contribute to sex-differential prevalence.

The paper introduces a framework to infer de novo exonic variants without parental genotypes, enhancing autism association studies by improving gene-based tests and controlling false discovery rates.

The paper introduces causarray, a causal inference framework for analyzing genomic data under unmeasured confounders, and applies it to identify causal effects in autism and Alzheimer's disease studies.

This paper introduces dbDiffusion, a diffusion-based framework for predicting transcriptional responses to unmeasured genetic perturbations using Perturb-seq data, achieving better accuracy than existing methods.

This study identifies rare coding variants contributing to OCD risk and shows genetic overlap with CTD and neurodevelopmental disorders, while showing minimal overlap with bipolar disorder and schizophrenia.

DAWN-SCAPE distinguishes between genes causing phenotypic differences and those affected by the phenotype, applying it to autism and Alzheimer's data to identify upstream and downstream biological processes.

The paper introduces XConTest, a method to assess genetic convergence in CRISPR perturbations, showing it identifies genes with similar downstream effects in autism and immune response studies.

The paper introduces a statistical framework for proteomic data analysis that combines imputation with debiasing to improve inference and reduce false positives in both single-cell and bulk data.

This paper identifies 35 genome-wide significant ASD risk genes in admixed Latin American populations, showing consistency with European cohorts and supporting the universality of ASD biology.

This study benchmarks association testing methods for low MOI perturb-seq data, identifies analysis challenges, and introduces SCEPTRE low-MOI to improve calibration and power.

eSVD-DE identifies differentially expressed genes across large cohorts of single-cell RNA-seq data by reducing dimensionality and accounting for individual-level confounding factors.

This paper improves the portability and fairness of polygenic scores by introducing Joint-Lassosum, which outperforms existing methods in diverse ancestries.

The paper identifies target genes and pathways of blood trait loci using CRISPR screens and single-cell RNA sequencing, revealing cis and trans regulatory effects of noncoding variants.

The paper constructs a cross-ancestry, cell-type-informed gene regulatory atlas of the developing human brain, identifying genetic regulation and linking it to neuropsychiatric disorders.

This paper introduces scVAEIT, a probabilistic variational autoencoder for integrating and imputing multimodal single-cell data with mosaic measurements, showing improved performance across unseen cell types and tissues.

The paper examines RNA editing changes during human cortical development, identifying expansion of editing sites and their genetic regulation linked to neurological disorders.

The paper introduces a model-free prediction test to assess if X contributes to predicting Y, showing improved power in gene expression analyses using CITE-seq and spatial transcriptomics data.

The paper introduces SCEPTRE, a method for analyzing single-cell CRISPR screens that improves calibration and sensitivity by addressing technical confounders through conditional resampling.

The paper develops a Bayesian method to estimate cell-type-specific gene expression from bulk RNA-seq data using priors from scRNA-seq, improving accuracy and enabling downstream analyses like eQTL discovery.

This paper presents a data harmonization pipeline for GWAS that improves power by integrating external controls and reduces batch effects, demonstrating its effectiveness in a Crohn's disease study.

This paper introduces a gene-based testing method that accounts for correlation between nearby genes in GWAS, improving power for detecting weak signals in autism spectrum disorder.

The paper identifies cell-type-specific marker genes from bulk transcriptome data by refining published markers using correlation patterns.

ESCO simulates single-cell gene expression with co-expression networks, showing imputation improves GCN recovery except when data are too sparse.

This paper introduces eSVD, a non-linear embedding method for single-cell RNA-seq data, and applies it to identify two distinct developmental trajectories in mouse oligodendrocytes.