Hildreth Robert Frost

LRSPAT: A low-rank framework for spatial omics statistics

LRSPAT is a low-rank framework for spatial omics statistics that improves speed, memory efficiency, and accuracy in identifying spatially variable genes in high-resolution spatial transcriptomics data.

Biological foundation models illuminate annotation blind spots in evolutionarily divergent genomes

The paper introduces a hybrid annotation framework that improves cross-species protein similarity assignment and gene annotation in evolutionarily divergent genomes, expanding structural catalogs and improving single-cell immune cell state annotation in sea lamprey.

Benchmarking sketching methods on spatial transcriptomics data

The paper benchmarks sketching methods on spatial transcriptomics data, showing that spatially aware methods outperform others in preserving tissue structure and capturing rare cell states.

Analysis of biological networks using Krylov subspace trajectories

The paper introduces a method using Krylov subspace trajectories for analyzing biological networks, demonstrating its application in community detection and perturbation analysis on the C. elegans neural network.

Parsimonious cell co-localization scoring for spatial transcriptomics

The paper introduces the Neighborhood Product Co-localization (NPC) score to quantify co-localization of cell types in spatial transcriptomics data, demonstrating its application on a mouse ovary dataset.

Autoencoder Denoising for Network-Based Spatial Transcriptomics Data with Applications for Cell Signaling Estimation

This paper introduces an autoencoder framework for denoising spatial transcriptomics networks and estimating cell signaling, showing improved performance over SVD on simulated data and biologically plausible results on real data.

Randomized Spatial PCA (RASP): a computationally efficient method for dimensionality reduction of high-resolution spatial transcriptomics data

RASP is a fast dimensionality reduction method for spatial transcriptomics data that improves computational efficiency while maintaining accuracy in spatial domain detection.

Gene set optimization for cancer transcriptomics using sparse principal component analysis

This paper optimizes gene set collections for cancer transcriptomics using sparse PCA, improving survival associations across 21 solid cancers.

Leveraging cell type-specificity for gene set analysis of single cell transcriptomics

The paper customizes gene set collections for single-cell RNA-seq by incorporating cell type-specific gene expression data to improve gene set testing power and interpretability.

Reconstruction Set Test (RESET): a computationally efficient method for single sample gene set testing based on randomized reduced rank reconstruction error

RESET is a single sample gene set testing method that improves accuracy and computational efficiency for detecting differential abundance and correlation patterns in scRNA-seq data.

Single cell transcriptomics-level Cytokine Activity Prediction and Estimation (SCAPE)

SCAPE predicts cell-level cytokine activity from scRNA-seq data using gene sets and a modified VAM method, validated on simulated and real datasets.

STREAK: A Supervised Cell Surface Receptor Abundance Estimation Strategy for Single Cell RNA-Sequencing Data using Feature Selection and Thresholded Gene Set Scoring

STREAK improves receptor abundance estimation in scRNA-seq by using supervised learning and gene set scoring, outperforming existing methods on human tissue datasets.

Cell type-specific Interaction Analysis using Doublets in scRNA-seq (CIcADA)

CIcADA identifies and analyzes biological doublets in scRNA-seq data, finding they upregulate immune response genes in tumor datasets.

SPECK: An Unsupervised Learning Approach for Cell Surface Receptor Abundance Estimation for Single Cell RNA-Sequencing Data

SPECK estimates cell surface receptor abundance from scRNA-seq data using an unsupervised approach, outperforming other methods on human receptors and multiple tissues.

CAMML with the Integration of Marker Proteins (ChIMP)

The paper introduces ChIMP, a method that improves cell typing for scRNA-seq and CITE-seq data by integrating transcriptomic and protein marker information.

A quaternion model for single cell transcriptomics

This paper introduces a quaternion-based model for scRNA-seq and spatial transcriptomics data to enable novel visualization, transformation, and spectral analysis of cell states and spatial patterns.

CBEA: Competitive balances for taxonomic enrichmentanalysis

CBEA provides a new method for taxonomic enrichment analysis in microbiome data by using a log-ratio formulation to generate sample-specific enrichment scores.

Evaluating trait-based sets for taxonomic enrichment analysis applied to human microbiome data sets

The paper evaluates trait-based taxonomic enrichment methods for human microbiome data, finding they improve disease prediction and interpretability.

Internal oligo(dT) priming in bulk and single cell RNA sequencing

The paper identifies and mitigates the effects of internal oligo(dT) priming in bulk and single cell RNA sequencing, improving gene expression measurement.

Tissue-adjusted pathway analysis of cancer (TPAC)

TPAC improves cancer prognosis and tumor stage prediction by analyzing pathway dysregulation in tumor transcriptomics data using normal tissue-specific gene information.

Variance-adjusted Mahalanobis (VAM): a fast and accurate method for cell-specific gene set scoring

The paper introduces VAM, a gene set testing method for scRNA-seq data that improves accuracy and efficiency compared to existing approaches.