Julia Salzman

The paper identifies a novel SARS-CoV-2-derived circular RNA, circ7b8N, which modulates host gene expression and is detectable in clinical samples, suggesting a role in host-pathogen interactions.

FLASH predicts phenotype from sequence using deep learning, achieving high accuracy on unseen variants and identifying new virulence predictors across microbes.

The paper introduces SPLASH, a reference-free algorithm for detecting sequence variation from RNA-seq data, and applies it to discover novel regulatory mechanisms in plant transcriptomes.

The paper benchmarks large language models for cell-type and gene set annotation using AnnDictionary, finding high accuracy in cell-type annotation and strong functional gene set recovery.

sc-SPLASH enables reference-free discovery in scRNA-seq and spatial transcriptomics, uncovering novel transcriptomic features like V(D)J recombination and secreted repeat proteins.

The paper develops a statistical framework, SPRAWL, to detect subcellular RNA localization and alternative 3' UTR regulation in mouse brain and liver cells, finding correlations between 3' UTR length and localization.

STRUCT identifies RNA secondary structures from raw sequencing data without multiple sequence alignment, rediscovering known elements and finding novel viral structures.

OASIS provides a statistically valid and computationally efficient alternative to Pearson's X2 for analyzing contingency tables, demonstrating improved power and interpretability in genomic applications.

The paper introduces SPLASH and compactors to efficiently discover microbial genetic elements and perform strain typing without reference genomes.

SPLASH is a unifying, statistically driven approach to biological discovery from raw sequencing data, bypassing alignment.

DIVE is a reference-free method for discovering mobile genetic elements and diversity-generating elements, showing improved detection power over existing reference-based approaches.

The paper was withdrawn due to duplicate posting and is no longer available for citation.

The paper benchmarks 16 circRNA detection tools across three human cell types, finding large differences in sensitivity but consistent precision, and recommends tool combinations for improved detection.

SPLASH2 is an efficient, scalable algorithm for discovering regulated sequence variation in RNA-seq data, revealing new biology in single-cell and bulk datasets without parameter tuning.

The paper applies RNA processing analysis methods to spatial transcriptomics data, identifying new spatially regulated RNA processing events in mouse tissues.

SPLASH+ identifies novel transcriptomic diversity in human single cells without reference genomes or cell type metadata.

ReadZS identifies cell type-specific and developmentally regulated RNA processing programs in single-cell RNA-seq without relying on pre-existing annotations.

This paper uses a mouse lemur transcriptomic atlas to identify new genes, splice junctions, and disease-related features, highlighting their relevance to primate biology and human disease.

Nearly 500,000 cells from 24 human tissues, many from one donor, characterize over 400 cell types and enable within-donor study of T-cell clones, B-cell mutation rates and splicing.

This paper identifies activated interstitial macrophages as a key target of SARS-CoV-2 infection in human lung tissue, showing their role in viral takeover, inflammation, and fibrosis.

The paper introduces SpliZ and SpliZVD, statistical methods to detect regulated splicing in single-cell RNA-seq, revealing new cell-type-specific splicing events in human lung cells.

The paper introduces SpliZ, a statistical method to detect cell-type-specific splicing regulation in single-cell RNA-seq data, finding 9.1% of genes show such regulation in human tissues.

SICILIAN improves splice junction detection in scRNA-seq by assigning statistical confidence, enhancing accuracy and agreement with bulk data and biological replicates.

This study shows TGS1 is essential for snRNA 3' end processing and its loss causes neurodegeneration, which can be mitigated by human TGS1 expression in model organisms.

This paper develops a high-throughput method to generate host and pathogen genomic data from single nasopharyngeal swabs of COVID-19 patients.

This paper introduces a new sampling theory, MEG, and its implementation SQUICH, which reduces sequencing depth requirements while maintaining precision.

DEEPEST improves gene fusion detection in cancer by reducing false positives and identifying new oncogenic RNA drivers using statistical methods.

The paper identifies cis-regulatory elements by analyzing gene trap insertions in haploid human cells using WNT signaling screens.

The paper identifies the promoter of ciRS-7 as overlapping with LINC00632 and shows ciRS-7 exonic sequence is embedded in linear transcripts in human and mouse.

This study develops a new algorithm to detect gene fusions in cancer genomes, identifying new recurrent fusions and candidate oncogenes, including in ovarian cancer.