Zachary B. Abrams

The paper corrects batch effects in the Answer ALS proteomics dataset using BDCs and ComBat, improving data utility while preserving biological variability.

The paper shows that pooled correlations in transcriptomic data often reverse sign when analyzed within specific contexts, highlighting the need for context-aware reporting of correlations.

The paper develops a method called Newmanization to reduce tissue signals in transcriptomic data and applies it to digestive tract cancer samples, showing improved clustering and separation compared to mutation data.

WayFindR converts pathway data into graph structures to analyze feedback loops, revealing their underrepresentation in curated databases.

This study evaluates 16 dimensionality reduction methods for CyTOF, scRNA-seq, and CITE-seq data, focusing on their ability to preserve pairwise distances and local neighborhoods.

This study identifies a small set of transcription factors and microRNAs that can predict genome-wide gene expression across multiple cancer types with high accuracy.

This study uses unsupervised clustering on longitudinal cognitive data to identify subtypes of dementia, finding that mild dementia subtypes are more heterogeneous in progression.

SillyPutty improves clustering by optimizing silhouette width, showing comparable accuracy to existing methods and better performance when combined with hierarchical clustering.

This paper compares GPT-4, GPT-3.5-turbo, Flan-T5, and spaCy-based methods for identifying cancer phenotypes from EHRs, finding GPT-4 achieves higher precision, recall, and F1 scores.

RCytoGPS converts JSON karyotype data into R objects for analysis and visualization, streamlining computational cytogenetic pathology.

The paper integrates mouse and human single-cell RNA sequencing data to predict and spatially map cell-type composition in human brain tissue.

The paper identifies genes differentially regulated by transcription factors and miRNAs using TCGA data and linear models to compare their regulatory effects.

Mercator is an R package that visualizes distance matrices using multiple algorithms to help researchers compare different distance metrics in unsupervised data analysis.

CytoGPS converts ISCN karyotype data into a machine-readable format and binary LGF model for downstream computational analysis.

Malachite automates gene enrichment meta-analyses across multiple datasets using ToppGene, identifying xenobiotic stimulus and NADP activity as enriched processes in smoking-related gene expression data.

Polychrome creates and evaluates color palettes with at least 24 distinguishable colors for categorical data visualization, addressing limitations in existing R packages.