Biology subjects

Kendziorski, C.

Publications and source records attributed to Kendziorski, C..

At least 19 recordsLinked to original sources

bioRxiv2026

Agentic Discovery Transfer Enables Cross-Cancer Discovery at Scale

Cancer research generates tens of thousands of studies each year, yet whether a published finding holds in another cancer type or patient population is rarely tested, because doing so is labor-intensive and difficult to scale. We define this task as discovery transfer and present SMART, a multi-agent large language model framework for discovery transfer in single-cell RNA-sequencing (scRNA-seq) studies. Given a source manuscript and a target scRNA-seq dataset, SMART extracts major clinical and mechanistic findings, generates and executes analyses for the testable findings, and reports which findings transfer and suggests follow-up. Applied to 121 studies across six cancer types, SMART identified findings that transfer across cancers, including a lung cancer prognostic signature that we confirmed to be associated with survival in liver cancer, pancreatic cancer and pleural mesothelioma. SMART also generated new mechanistic hypotheses, demonstrating how agentic discovery transfer can extend published findings beyond their original setting.

bioinformatics↗

bioRxiv2026

Spatial multi-omics and single-cell transcriptomics uncover senescence-associated cellular programs during colon adenoma to cancer progression

Colorectal cancer develops through a normal-adenoma-carcinoma sequence, yet only 5-10% of adenomas progress to malignancy, and the cellular programs governing that sequence remain poorly defined. Here we generate a spatial multi-omics atlas of human colon adenomas, combining Visium CytAssist and protein co-detection across 24 nonadvanced and advanced tubular adenomas with single-cell resolution Xenium Prime 5K profiling of 101 patient-matched normal, adenoma, and carcinoma cores from 16 patients. Integrating whole-transcriptome and 31-plex protein data identifies nine spatial clusters and two dysplastic epithelial populations that co-express stemness, proliferation, and senescence programs. These programs occupy a shared, spatially confined epithelial niche that expands from adenoma to carcinoma. Spatial analysis revealed GDF15, a senescence-associated secretory factor, mediated the coupling between senescence and stemness in advanced adenomas, and that GDF15-high epithelium locally excludes CD8+ T cells in adenoma and, more broadly, in carcinoma. These findings position senescence as a spatially instructive rather than merely tumor-suppressive program during colorectal carcinogenesis and suggest GDF15 might be a potential candidate target for cancer prevention and interception in the colon.

cancer biology↗

bioRxiv2026

AutoZyme: An Autonomous Agentic Framework to Optimize Bioinformatics Software

Performance bottlenecks in widely used genomics and bioinformatics software present a substantial and growing burden as biological datasets continue to increase in size and number. Relieving these bottlenecks relies largely on expert manual optimization and therefore remains difficult to scale. Here we present AutoZyme, an agentic framework for scientific software optimization. Given a target function, AutoZyme builds benchmarks, identifies bottlenecks, and iteratively tests code changes, retaining only those that improve runtime while preserving output. We evaluated AutoZyme on 45 functions, improving runtime without substantial memory increases in over 95% of cases considered. Across 38 functions from Seurat, Scanpy and related packages in genomics and bioinformatics, AutoZyme reduced runtime by a median of 8.52-fold, with the largest reductions exceeding 676-fold. The optimized functions are distributed through AutoZyme-Library as drop-in replacements for existing analysis pipelines. We also release AutoZyme as a reusable framework for optimizing additional user-specified packages and functions.

bioinformatics↗

bioRxiv2024

CASSIA allows for robust, automated cell annotation in single-cell RNA-sequencing data

Cell type annotation is an essential step in single-cell RNA-sequencing analysis, and numerous annotation methods are available. Most require a combination of computational and domain-specific expertise, and they frequently yield inconsistent results that can be challenging to interpret. Large language models have the potential to expand accessibility while reducing manual input and improving accuracy, but existing approaches suffer from hyperconfidence, hallucinations, and lack of reasoning. To address these limitations, we developed CASSIA for automated, accurate, and interpretable cell annotation of single-cell RNA-sequencing data. As demonstrated in analyses of over 970 cell types, CASSIA improves annotation accuracy in benchmark datasets as well as complex and rare cell populations, and also provides users with reasoning and quality assessment to ensure interpretability, guard against hallucinations, and calibrate confidence.

bioinformatics↗

bioRxiv2023

Targeted inhibition of BET proteins in HPV-16 associated head and neck squamous cell carcinoma reveals heterogeneous transcription response

Integrated human papillomavirus (HPV-16) associated head and neck squamous cell carcinoma (HNSCC) tumors have worse survival outcomes compared to episomal HPV-16 HNSCC tumors. Therefore, there is a need to differentiate treatment for HPV-16 integrated HNSCC from other viral forms. We analyzed TCGA data and found that HPV+ HNSCC expressed higher transcript levels of the bromodomain and extra terminal domain (BET) family of transcriptional coregulators. However, the mechanism of BET protein-mediated transcription of viral-cellular genes in the integrated viral-HNSCC genomes needs to be better understood. We show that BET inhibition downregulates E6 significantly independent of the viral transcription factor, E2, and there was overall heterogeneity in the downregulation of viral transcription in response to the effects of BET inhibition across HPV-associated cell lines. Chemical BET inhibition was phenocopied with the knockdown of BRD4 and mirrored downregulation of viral E6 and E7 expression. Strikingly, there was heterogeneity in the reactivation of p53 levels despite E6 downregulation, while E7 downregulation did not alter Rb levels significantly. We identified that BET inhibition directly downregulated c-Myc and E2F expression and induced CDKN1A expression. Overall, our studies show that BET inhibition provokes a G1-cell cycle arrest with apoptotic activity and suggests that BET inhibition regulates both viral and cellular gene expression in HPV-associated HNSCC.

cancer biology↗

bioRxiv2023

SpatialView: An interactive web application forvisualization of multiple samples in spatialtranscriptomics experiments

Spatial transcriptomics (ST) experiments provide spatially localized measurements of genome-wide gene expression allowing for an unprecedented opportunity to investigate cellular heterogeneity and organization within a tissue. Statistical and computational frameworks exist that implement robust methods for pre-processing and analyzing data in ST experiments. However, the lack of an interactive suite of tools for visualizing ST data and results currently limits the full potential of ST experiments. To fill this gap, we developed SpatialView, an open-source web browser-based interactive application for visualizing data and results from multiple 10x Genomics ST experiments. We anticipate SpatialView will be useful to a broad array of clinical and basic science investigators utilizing ST to study disease. AvailabilitySpatialView is available at https://github.com/kendziorski-lab/SpatialView ; a demo application is available at https://www.biostat.wisc.edu/~kendzior/spatialviewdemo/ Contactcmohanty2@wisc.edu, ckendziorski@wisc.edu Supplementary informationSupplementary data are available at Bioinformatics online.

bioinformatics↗

bioRxiv2022

SpatialCorr: Identifying Gene Sets with Spatially Varying Correlation Structure

Recent advances in spatially resolved transcriptomics technologies enable both the measurement of genome-wide gene expression profiles and their mapping to spatial locations within a tissue. A first step in spatial transcriptomics data analysis is identifying genes with expression that varies spatially, and robust statistical methods exist to address this challenge. While useful, these methods do not detect spatial changes in the coordinated expression within a group of genes. To this end, we present SpatialCorr, a method for identifying sets of genes with spatially varying correlation structure. Given a collection of gene sets pre-defined by a user, SpatialCorr tests for spatially induced differences in the correlation of each gene set within tissue regions, as well as between and among regions. An application to cutaneous squamous cell carcinoma demonstrates the power of the approach for revealing biological insights not identified using existing methods.

bioinformatics↗

bioRxiv2021

SpotClean adjusts for spot swapping in spatial transcriptomics data

Spatial transcriptomics (ST) is a powerful and widely-used approach for profiling genome-wide gene expression across a tissue with emerging applications in molecular medicine and tumor diagnostics. Recent spatial transcriptomics experiments utilize slides containing thousands of spots with spot-specific barcodes that bind mRNA. Ideally, unique molecular identifiers at a spot measure spot-specific expression, but this is often not the case owing to bleed from nearby spots, an artifact we refer to as spot swapping. We propose SpotClean to adjust for spot swapping and, in doing so, to increase the sensitivity and precision with which downstream analyses are conducted.

genomics↗

bioRxiv2021

The Anna Karenina model of beta cell maturation in development and their dedifferentiation in type 1 and type 2 diabetes

Loss of mature {beta} cell function and identity, or {beta} cell dedifferentiation, is seen in all types of diabetes mellitus. Two competing models explain {beta} cell dedifferentiation in diabetes. In the first model, {beta} cells dedifferentiate in the reverse order of their developmental ontogeny. This model predicts that dedifferentiated {beta} cells resemble {beta} cell progenitors. In the second model, {beta} cell dedifferentiation depends on the type of diabetogenic stress. This model, which we call the "Anna Karenina" model, predicts that in each type of diabetes, {beta} cells dedifferentiate in their own way, depending on how their mature identity is disrupted by any particular diabetogenic stress. We directly tested the two models using a {beta} cell-specific lineage-tracing system coupled with RNA-sequencing in mice. We constructed a multidimensional map of {beta} cell transcriptional trajectories during the normal course of {beta} cell postnatal development and during their dedifferentiation in models of both type 1 diabetes (NOD) and type 2 diabetes (BTBR-Lepob/ob). Using this unbiased approach, we show here that despite some similarities between immature and dedifferentiated {beta} cells, {beta} cells dedifferentiation in the two mouse models is not a reversal of developmental ontogeny and is different between different types of diabetes.

developmental biology↗

bioRxiv2020

Normalization by distributional resampling of high throughput single-cell RNA-sequencing data

MotivationNormalization to remove technical or experimental artifacts is critical in the analysis of single-cell RNA-sequencing experiments, even those for which unique molecular identifiers (UMIs) are available. The majority of methods for normalizing single-cell RNA-sequencing data adjust average expression in sequencing depth, but allow the variance and other properties of the gene-specific expression distribution to be non-constant in depth, which often results in reduced power and increased false discoveries in downstream analyses. This problem is exacerbated by the high proportion of zeros present in most datasets. ResultsTo address this, we present Dino, a normalization method based on a flexible negative-binomial mixture model of gene expression. As demonstrated in both simulated and case study datasets, by normalizing the entire gene expression distribution, Dino is robust to shallow sequencing depth, sample heterogeneity, and varying zero proportions, leading to improved performance in downstream analyses in a number of settings. Availability and implementationThe R package, Dino, is available on GitHub at https://github.com/JBrownBiostat/Dino. Contactbrownj@biostat.wisc.edu, kendzior@biostat.wisc.edu

genomics↗

bioRxiv2020

Enhancing biological signals and detection rates in single-cell RNA-seq experiments with cDNA library equalization

Considerable effort has been devoted to refining experimental protocols having reduced levels of technical variability and artifacts in single-cell RNA-sequencing data (scRNA-seq). We here present evidence that equalizing the concentration of cDNA libraries prior to pooling, a step not consistently performed in single-cell experiments, improves gene detection rates, enhances biological signals, and reduces technical artifacts in scRNA-seq data. To evaluate the effect of equalization on various protocols, we developed Scaffold, a simulation framework that models each step of an scRNA-seq experiment. Numerical experiments demonstrate that equalization reduces variation in sequencing depth and gene-specific expression variability. We then performed a set of experiments in vitro with and without the equalization step and found that equalization increases the number of genes that are detected in every cell by 17-31%, improves discovery of biologically relevant genes, and reduces nuisance signals associated with cell cycle. Further support is provided in an analysis of publicly available data.

genomics↗

bioRxiv2020

CHARTS: A web application for characterizing and comparing tumor subpopulations in publicly available single-cell RNA-seq datasets

BackgroundSingle-cell RNA-seq (scRNA-seq) enables the profiling of genome-wide gene expression at the single-cell level and in so doing facilitates insight into and information about cellular heterogeneity within a tissue. Perhaps nowhere is this more important than in cancer, where tumor and tumor microenvironment heterogeneity directly impact development, maintenance, and progression of disease. While publicly available scRNA-seq cancer datasets offer unprecedented opportunity to better understand the mechanisms underlying tumor progression, metastasis, drug resistance, and immune evasion, much of the available information has been underutilized, in part, due to the lack of tools available for aggregating and analysing these data. ResultsWe present CHARacterizing Tumor Subpopulations (CHARTS), a computational pipeline and web application for analyzing, characterizing, and integrating publicly available scRNA-seq cancer datasets. CHARTS enables the exploration of individual gene expression, cell type, malignancy-status, differentially expressed genes, and gene set enrichment results in subpopulations of cells across multiple tumors and datasets. ConclusionCHARTS is an easy to use, comprehensive platform for exploring single-cell subpopulations within tumors across the ever-growing collection of public scRNA-seq cancer datasets. CHARTS is freely available at charts.morgridge.org.

bioinformatics↗

bioRxiv2020

Interspecies Chimeric Conditions Affect the Developmental Rate of Human Pluripotent Stem Cells

Human pluripotent stem cells hold significant promise for regenerative medicine. However, long differentiation protocols and immature characteristics of stem cell-derived cell types remain challenges to the development of many therapeutic applications. In contrast to the slow differentiation of human stem cells in vitro that mirrors a nine-month gestation period, mouse stem cells develop according to a much faster three-week gestation timeline. Here, we tested if co-differentiation with mouse pluripotent stem cells could accelerate the differentiation speed of human embryonic stem cells. Following a six-week RNA-sequencing time course of neural differentiation, we identified 929 human genes that were upregulated earlier and 535 genes that exhibited earlier peaked expression profiles in chimeric cell cultures than in human cell cultures alone. Genes with accelerated upregulation were significantly enriched in Gene Ontology terms associated with neurogenesis, neuron differentiation and maturation, and synapse signaling. Moreover, chimeric mixed samples correlated with in utero human embryonic samples earlier than human cells alone, and acceleration was dose-dependent on human-mouse co-culture ratios. Differences in the timing and expression levels of genes corresponding to neuron cell types and brain region identity under chimeric conditions were also observed. The altered developmental rates and lineage outcomes described in this report have implications for accelerating human stem cell differentiation and the use of interspecies chimeric embryos in developing human organs for transplantation. Author SummaryHuman pluripotent stem cells often require long in vitro protocols to form mature cell types of clinical relevance for potential regenerative therapies, a ramification of a nine-month developmental clock in utero that also runs ex utero. What controls species-specific developmental time and whether the timer is amenable to acceleration is unknown. Further, interspecies chimeric embryos are increasingly being created to study early human development or explore the potential growth of human organs for transplantation. How the conflicting developmental speeds of cells from different species co-differentiating together affect each other is not understood. Here, using genome-wide transcriptional analysis of RNA-sequencing time courses, we show that 1) co-differentiating human embryonic stem cells intermixed with mouse stem cells accelerated elements of human developmental programs, 2) the acceleration was dose-dependent on the proportion of mouse cells, and 3) human cells in chimeric samples correlated to in utero samples earlier than human only samples. Our results provide evidence that some components of species-specific developmental clocks may be susceptible to acceleration.

developmental biology↗

bioRxiv2020

EBSeq: improving mixing computations for multi-group differential expression analysis

EBSeq is a Bioconductor package designed to calculate empirical-Bayesian inference summaries from sequence-based gene-expression (RNA-Seq) data. It produces gene or isoform-specific scores that measure various patterns of differential expression among a set of sample groups, and is most commonly deployed to measure differential expression between two groups. Its use of local posterior probabilities from a fitted mixture model provides the data analyst a direct way to score the false discovery rate of any reported list of genes, and it is one of the only tools that can address local false discovery rates when analyzing multiple sample groups. Contemporary applications have increasing numbers of sample groups, and the algorithms deployed in EBSeq are neither space nor time efficient in this important case. We describe a version update utilizing code improvements and novel pruning and clustering algorithms in order to reduce the complexity of mixture computations. The algorithms are supported by a theoretical analysis and tested empirically on a variety of benchmark and synthetic data sets.

bioinformatics↗

bioRxiv2019

CB2 distinguishes cells from background barcodes in 10x Genomics data

An important challenge in pre-processing data from the 10x Genomics Chromium platform is distinguishing barcodes associated with real cells from those binding background reads. Existing methods test barcodes individually, and consequently do not leverage the strong cell-to-cell correlation present in most datasets. To improve the power to identify real cells and rare subpopulations, we introduce CB2, a cluster-based approach for distinguishing real cells from background barcodes.

genomics↗

bioRxiv2019

Tradeoff between more cells and higher read depth for single-cell RNA-seq spatial ordering analysis of the liver lobule

As newer single-cell protocols generate increasingly more cells at reduced sequencing depths, the value of a higher read depth may be overlooked. Using data from three different single-cell RNA-seq protocols that lend themselves to having either higher read depth (Smart-seq) or many cells (MARS-seq and 10X), we evaluate their ability to recapitulate biological signals in the context of pseudo-spatial reconstruction. Overall, we find gene expression profiles after spatial-reconstruction analysis are highly reproducible between datasets despite being generated by different protocols and using different computational algorithms. While UMI based protocols such as 10X and MARS-seq allow for capturing more cells, Smart-seqs higher sensitivity and read-depth allows for analysis of lower expressed genes and isoforms. Additionally, we evaluate trade-offs for each protocol by performing subsampling analyses, and find that optimizing the balance between sequencing depth and number of cells within a protocol is important for efficient use of resources. Our analysis emphasizes the importance of selecting a protocol based on the biological questions and features of interest.

genomics↗

bioRxiv2019

A Compositional Model To Assess Expression Changes From Single-Cell RNA-Seq Data

On the problem of scoring genes for evidence of changes in the distribution of single-cell expression, we introduce an empirical Bayesian mixture approach and evaluate its operating characteristics in a range of numerical experiments. The proposed approach leverages cell-subtype structure revealed in cluster analysis in order to boost gene-level information on expression changes. Cell clustering informs gene-level analysis through a specially-constructed prior distribution over pairs of multinomial probability vectors; this prior meshes with available model-based tools that score patterns of differential expression over multiple subtypes. We derive an explicit formula for the posterior probability that a gene has the same distribution in two cellular conditions, allowing for a gene-specific mixture over subtypes in each condition. Advantage is gained by the compositional structure of the model, in which a host of gene-specific mixture components are allowed, but also in which the mixing proportions are constrained at the whole cell level. This structure leads to a novel form of information sharing through which the cell-clustering results support gene-level scoring of differential distribution. The result, according to our numerical experiments, is improved sensitivity compared to several standard approaches for detecting distributional expression changes.

bioinformatics↗

bioRxiv2017

Trendy: Segmented regression analysis of expression dynamics for high-throughput ordered profiling experiments

AbstractHigh throughput expression profiling experiments with ordered conditions (e.g. time-course or spatial-course) are becoming more common for profiling detailed differentiation processes or spatial patterns. Identifying dynamic changes at both the individual gene and whole transcriptome level can provide important insights about genes, pathways, and critical time-points. We present an R package, Trendy, which utilizes segmented regression models to simultaneously characterize each genes expression pattern and summarize overall dynamic activity in ordered condition experiments. For each gene, Trendy finds the optimal segmented regression model and provides the location and direction of dynamic changes in expression. We demonstrate the utility of Trendy to provide biologically relevant results on both microarray and RNA-seq datasets. Trendy is a flexible R package which characterizes gene-specific expression patterns and summarizes changes of global dynamics over ordered conditions. Trendy is freely available as an R package with a full vignette at https://github.com/rhondabacher/Trendy.

genomics↗