Biology subjects

Niu, S.-Y.

Publications and source records attributed to Niu, S.-Y..

6 recordsLinked to original sources

bioRxiv2026

Angiopoietin-like protein 7 (ANGPTL7)-mediated effects on resident immune cells contribute to neurodegeneration in glaucomatous optic nerve head

Glaucoma is an optic neuropathy characterized by progressive retinal degeneration. Population genetic analyses describe an association of variants in angiopoietin-like protein 7 (ANGPTL7) with lower intraocular pressure (IOP) and decreased glaucoma susceptibility. In addition to its reported role in the trabecular meshwork (TM), we demonstrate that ANGPTL7 is also expressed within disease-relevant cell types of the human posterior eye, specifically prelaminar optic nerve head astrocytes (ONHA). Contrasting with previously published findings, we observe no difference in IOP between Angptl7-/- mice and wild-type littermates. Despite absence of an IOP phenotype, loss of ANGPTL7 protects from IOPinduced retinal ganglion cell (RGC) damage in a rodent glaucoma neurodegeneration model. ANGPTL7 may signal through LAIR1-positive immune cells, potentially contributing to pro-inflammatory signaling and phenotypic modulation of resident immune cells, with evidence suggesting that oligomerization of ANGPTL7 is necessary for its activity. Comprehensively, our data highlight an unappreciated role of ANGPTL7 in modulating neuroinflammation in the posterior eye, contributing to glaucoma progression.

neuroscience↗

bioRxiv2025

Generative genomics accurately predicts future experimental results

Realizing AIs promise to accelerate biomedical research requires AI models that are both accurate and sufficiently flexible to capture the diversity of real-life experiments. Here, we describe a generative genomics framework for AI-based experimental prediction that mirrors the process of designing and conducting an experiment in the lab or clinic. We created GEM-1 (Generate Expression Model-1), an AI system that effectively models the enormous range of bulk and single-cell gene expression experiments performed by scientists and benchmarked its performance across multiple biological axes. GEM-1s prediction of future gene expression experiments-RNA-seq data deposited in public archives after our training data cutoff-yielded accuracy comparable to the best-possible performance estimated by comparing the results of matched lab experiments. Overall, our approach illustrates the transformative potential of generative genomics for applications ranging from predicting cellular perturbations in vitro to de novo generation of data from large clinical cohorts.

genomics↗

bioRxiv2020

Forkhead box R1-mediated stress response linked to a case of human microcephaly and brain atrophy

Forkhead box (Fox) family transcription factors are highly conserved and play essential roles in a wide range of cellular and developmental processes. We report an individual with severe neurological symptoms including postnatal microcephaly, progressive brain atrophy and global developmental delay associated with a de novo missense variant (M280L) in the FOXR1 gene. At the protein level, M280L impaired FOXR1 expression and induced a nuclear aggregate phenotype due to protein misfolding and proteolysis. RNAseq and pathway analysis showed that FOXR1 acts as both a transcriptional activator and repressor with central roles in heat shock response, chaperone cofactor-dependent protein refolding and cellular response to stress. Indeed, FOXR1 expression is increased in response to cellular stress, a process in which it directly controls HSPA6, HSPA1A and DHRS2 transcripts. Meanwhile, the ability of the M280L mutant to respond to stress is compromised, in part due to impaired regulation of downstream target genes that are involved in the stress response pathway. Combined, these results suggest FOXR1 plays a role in cellular stress and is necessary for normal brain development.

molecular biology↗

bioRxiv2020

Epigenomic Diversity of Cortical Projection Neurons in the Mouse Brain

Neuronal cell types are classically defined by their molecular properties, anatomy, and functions. While recent advances in single-cell genomics have led to high-resolution molecular characterization of cell type diversity in the brain, neuronal cell types are often studied out of the context of their anatomical properties. To better understand the relationship between molecular and anatomical features defining cortical neurons, we combined retrograde labeling with single-nucleus DNA methylation sequencing to link epigenomic properties of cell types to neuronal projections. We examined 11,827 single neocortical neurons from 63 cortico-cortical (CC) and cortico-subcortical long-distance projections. Our results revealed unique epigenetic signatures of projection neurons that correspond to their laminar and regional location and projection patterns. Based on their epigenomes, intra-telencephalic (IT) cells projecting to different cortical targets could be further distinguished, and some layer 5 neurons projecting to extra-telencephalic targets (L5-ET) formed separate subclusters that aligned with their axonal projections. Such separation varied between cortical areas, suggesting area-specific differences in L5-ET subtypes, which were further validated by anatomical studies. Interestingly, a population of CC projection neurons clustered with L5-ET rather than IT neurons, suggesting a population of L5-ET cortical neurons projecting to both targets (L5-ET+CC). We verified the existence of these neurons by labeling the axon terminals of CC projection neurons and observed clear labeling in ET targets including thalamus, superior colliculus, and pons. These findings highlight the power of single-cell epigenomic approaches to connect the molecular properties of neurons with their anatomical and projection properties.

neuroscience↗

bioRxiv2019

Single nucleus multi-omics links human cortical cell regulatory genome diversity to disease risk variants

Single-cell technologies enable measure of unique cellular signatures, but are typically limited to a single modality. Computational approaches allow integration of diverse single-cell datasets, but their efficacy is difficult to validate in the absence of authentic multi-omic measurements. To comprehensively assess the molecular phenotypes of single cells in tissues, we devised single-nucleus methylCytosine, Chromatin accessibility and Transcriptome sequencing (snmC2T-seq) and applied it to post-mortem human frontal cortex tissue. We developed a computational framework to validate fine-grained cell types using multi-modal information and assessed the effectiveness of computational integration methods. Correlation analysis in individual cells revealed distinct relations between methylation and gene expression. Our integrative approach enabled joint analyses of the methylome, transcriptome, chromatin accessibility and conformation for 63 human cortical cell types. We reconstructed regulatory lineages for cortical cell populations and found specific enrichment of genetic risk for neuropsychiatric traits, enabling prediction of cell types with causal roles in disease.

genomics↗

bioRxiv2019

rSeqTU -- a machine-learning based R package for prediction of bacterial transcription units

A transcription unit (TU) is composed of one or multiple adjacent genes on the same strand that are co-transcribed in mostly prokaryotes. Accurate identification of TUs is a crucial first step to delineate the transcriptional regulatory networks and elucidate the dynamic regulatory mechanisms encoded in various prokaryotic genomes. Many genomic features, e.g., gene intergenic distance, and transcriptomic features including continuous and stable RNA-seq reads count signals, have been collected from a large amount of experimental data and integrated into classification techniques to computationally predict genome-wide TUs. Although some tools and web servers are able to predict TUs based on bacterial RNA-seq data and genome sequences, there is a need to have an improved machine-learning prediction approach and a better comprehensive pipeline handling QC, TU prediction, and TU visualization. To enable users to efficiently perform TU identification on their local computers or high-performance clusters and provide a more accurate prediction, we develop an R package, named rSeqTU. rSeqTU uses a random forest algorithm to select essential features describing TUs and then uses support vector machine (SVM) to build TU prediction models. rSeqTU (available at https://s18692001.github.io/rSeqTU/) has six computational functionalities including read quality control, read mapping, training set generation, random-forest-based feature selection, TU prediction, and TU visualization.

bioinformatics↗