Biology subjects

Chen, Y. T.

Publications and source records attributed to Chen, Y. T..

9 recordsLinked to original sources

bioRxiv2026

Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Accurate predictions of transcriptomic responses to genetic perturbations could unlock our understanding of gene functions and regulatory networks. While a growing number of methods and benchmarks target this task, existing evaluations focus on mean expression accuracy alone. This overlooks differential expression (DE), which captures both mean and variance and forms the basis for biological interpretation and experimental follow-up. Here, we systematically evaluate a diverse set of deep learning and non-deep-learning methods for their ability to predict DE outcomes under two generalization regimes: unseen perturbations within the same cell line, and unseen cellular contexts across cell lines. We find that simple baselines, such as embedding-based nearest neighbors, are competitive and often outperform specialized deep learning models for DE classification across datasets and evaluation metrics. We further show that sparsity calibration, motivated by the structure of single-cell data, substantially improves DE classification for deep learning models that do not explicitly account for sparsity. Together, our findings establish practical baselines and evaluation principles for benchmarking perturbation models on DE prediction.

genomics↗

bioRxiv2026

Mind the Alignment Gap: A Spatial Transcriptomics Benchmark for Scientific Coding Agents

AO_SCPLOWBSTRACTC_SCPLOWScientific coding agents are difficult to benchmark because many research tasks require executable work yet produce ambiguous or hard-to-verify outputs. Because benchmark construction requires substantial time and resources, automation offers a path to accelerating methods evaluation. We introduce an interactive framework for constructing scientificagent benchmarks from peer-reviewed papers and diagnosing agent behavior through trace inspection. We apply it as a case study in spatial transcriptomics alignment, constructing 40 tasks from SABench in which agents submit coordinate tables aligning pairs of two-dimensional tissue slices. Across 120 runs and three configurations, we compare a basic prompt, a package-aware prompt, and a full prompt with a prebuilt virtual environment. In this setting, richer package and environment context increased tool exploration but reduced the mean alignment score relative to the basic prompt (0.36 vs. 0.43; 95% CI, [-0.11,-0.03]). Trace inspection showed that added scaffolding often induced unnecessary transformations, fragile package-first workflows, and infrastructure failures. These results illustrate how specialized tooling can alter agent behavior and why scientific-agent benchmarks should evaluate agent traces and the workflows that produce them in addition to the final outputs.

bioinformatics↗

bioRxiv2026

T6SS mutants exploit itaconate to support infection of phagocytes

Pseudomonas aeruginosa is a major cause of persistent pneumonias that are not readily cleared by seemingly appropriate antimicrobial therapy. We identified a reservoir of P. aeruginosa variants lacking expression of the H3-T6SS in patients with chronic but not acute pneumonia. A PAO1 {Delta}H3-T6 mutant caused increased infection in the murine lung as compared to the wild-type strain. The {Delta}H3 mutants exhibited increased transcription of genes involved in phagocytic uptake and respiration under conditions found in the phagolysosome, namely low O2, low pH and abundant itaconate. We confirmed increased intraphagocytic residence of the {Delta}H3 mutants and colocalization with LAMP1 within the phagolysosome of both bone marrow derived macrophages in vitro and in alveolar macrophages harvested directly from infected lungs. Persistence within macrophages required itaconate which preserved the viability of infected macrophages and boosted bacterial bioenergetics to optimize consumption of available carbon sources. Our findings demonstrate that selection for loss of H3-T6SS loss of function mutations promotes the metabolic versatility that enables P. aeruginosa to cause intractable pulmonary infection.

immunology↗

bioRxiv2025

Sequestration of the phagocyte metabolite itaconate by P. aeruginosa RpoN promotes successful pulmonary infection

AbstractThe phagocyte immunometabolite itaconate, normally toxic to bacteria, functions as a signal to stimulate the adaptation of the pulmonary pathogen Pseudomonas aeruginosa to the lung. Itaconate is actively transported into P. aeruginosa where it induces {sigma}54 rpoN expression and co-valently binds cysteine residues on RpoN. RpoN not only functions as a sink to limit itaconate toxicity but S- itaconated RpoN promotes increased utilization of the Entner Doudoroff pathway, optimizing bacterial metabolism in the setting of inflammation. S-itaconation of RpoN directs a global metabolic response that fuels pulmonary infection.

microbiology↗

bioRxiv2025

RL-Finetuning of OpenAI o1-mini to Enhance Biomedical Reasoning

Recent breakthroughs in advanced reasoning large language models (LLMs), such as OpenAIs o1, have achieved impressive results in domains like math and coding. However, its not clear how much this type of reasoning helps in solving biomedical problems that involve more domain specialized knowledge and open-ended reasoning. Across two biomedical domains--gene characterization and small molecule property prediction--we find that the commercially available o1-mini model does not consistently outperform non-reasoning LLMs like GPT-4o. This motivated us to explore how much we can improve o1-minis biomedical reasoning through reinforcement learning (RL) finetuning. We show that RL finetuning of o1-mini results in large improvements in performance on gene classification, where it surprisingly outperformed domain-specific state-of-the-art models on some tasks. The results are mixed for small molecule prediction, suggesting that chemical reasoning could be more challenging for LLMs. We conclude with a discussion of the challenges and takeaways from this initial exploration of RL finetuning reasoning models for biomedical tasks.

bioinformatics↗

bioRxiv2025

Chronic degenerative failure of salivary glands can be reversed through restoring mitochondrial function

Chronic degenerative wounds are often deemed irreparable, directing research efforts to focus predominantly on acute tissue injury regeneration while leaving endogenous repair mechanisms for chronically damaged tissues largely unexplored. In this study, we demonstrate that non-healing, severely degenerated salivary gland tissues can be fundamentally restored through first-line treatment with muscarinic agonists. This approach rescues tissue structure and function, returning it to a homeostatic-like state, and reactivates endogenous regeneration processes to drive new cell expansion that persists for months post-treatment. Furthermore, neuromimetic activation profoundly depletes radiation-induced DNA damage and re-establishes the nerve-acinar relationship, ultimately restoring the tissues physiological capacity to maintain homeostasis, even in the absence of treatment. We show that full recovery of organ function, comparable to uninjured controls, is primarily mediated by the re-differentiation of aberrantly de-differentiated epithelial acinar cells and the restoration of mitochondrial function via a muscarinic-calcium signaling pathway. These findings challenge the prevailing notion that chronic organ degeneration is irreversible and propose a readily testable therapeutic strategy for epithelial restoration with potential applications across a spectrum of chronic injuries.

cell biology↗

bioRxiv2024

GenePert: Leveraging GenePT Embeddings for Gene Perturbation Prediction

Predicting how perturbation of a target gene affects the expression of other genes is a critical component of understanding cell biology. This is a challenging prediction problem as the model must capture complex gene-gene relationships and the output is high-dimensional and sparse. To address this challenge, we present GenePert, a simple approach that leverages GenePT embeddings, which are derived using ChatGPT from text descriptions of individual genes, to predict gene expression changes due to perturbations via regularized regression models. Benchmarked on eight CRISPR perturbation screen datasets across multiple cell types and five different pretrained gene embedding models, GenePert consistently outperforms all the state-of-the-art prediction models measured in both Pearson correlation and mean squared error metrics. Even with limited training data, our model generalizes effectively, offering a scalable solution for predicting perturbation outcomes. These findings underscore the power of informative gene embeddings in predicting the outcomes of unseen genetic perturbation experiments in silico. GenePert is available at https://github.com/zou-group/GenePert.

bioinformatics↗

bioRxiv2023

GENEPT: A SIMPLE BUT HARD-TO-BEAT FOUNDATION MODEL FOR GENES AND CELLS BUILT FROM CHATGPT

There has been significant recent progress in leveraging large-scale gene expression data to develop foundation models for single-cell biology. Models such as Geneformer and scGPT implicitly learn gene and cellular functions from the gene expression profiles of millions of cells, which requires extensive data curation and resource-intensive training. Here we explore a much simpler alternative by leveraging ChatGPT embeddings of genes based on literature. Our proposal, GenePT, uses NCBI text descriptions of individual genes with GPT-3.5 to generate gene embeddings. From there, GenePT generates single-cell embeddings in two ways: (i) by averaging the gene embeddings, weighted by each genes expression level; or (ii) by creating a sentence embedding for each cell, using gene names ordered by the expression level. Without the need for dataset curation and additional pretraining, GenePT is efficient and easy to use. On many downstream tasks used to evaluate recent single-cell foundation models -- e.g., classifying gene properties and cell types -- GenePT achieves comparable, and often better, performance than Geneformer and other models. GenePT demonstrates that large language model embedding of literature is a simple and effective path for biological foundation models.

bioinformatics↗

bioRxiv2023

Microbes with higher metabolic independence are enriched in human gut microbiomes under stress

A wide variety of human diseases are associated with loss of microbial diversity in the human gut, inspiring a great interest in the diagnostic or therapeutic potential of the microbiota. However, the ecological forces that drive diversity reduction in disease states remain unclear, rendering it difficult to ascertain the role of the microbiota in disease emergence or severity. One hypothesis to explain this phenomenon is that microbial diversity is diminished as disease states select for microbial populations that are more fit to survive environmental stress caused by inflammation or other host factors. Here, we tested this hypothesis on a large scale, by developing a software framework to quantify the enrichment of microbial metabolisms in complex metagenomes as a function of microbial diversity. We applied this framework to over 400 gut metagenomes from individuals who are healthy or diagnosed with inflammatory bowel disease (IBD). We found that high metabolic independence (HMI) is a distinguishing characteristic of microbial communities associated with individuals diagnosed with IBD. A classifier we trained using the normalized copy numbers of 33 HMI-associated metabolic modules not only distinguished states of health versus IBD, but also tracked the recovery of the gut microbiome following antibiotic treatment, suggesting that HMI is a hallmark of microbial communities in stressed gut environments.

microbiology↗