Biology subjects

Hon, G.

Publications and source records attributed to Hon, G..

6 recordsLinked to original sources

bioRxiv2026

pertTF: context-aware AI modeling for genome-scale and cross-system perturbation prediction

Predicting genetic perturbation responses at a single-cell level is central to building models for cell state and disease. However, existing approaches are limited on predicting phenotypic outcomes beyond expression changes and generalizing predictions across genome-scale perturbations in biologically relevant contexts. Here we introduce pertTF, a transformer-based single-cell genetic perturbation model. pertTF was trained from a unique dataset capturing single cell expressions profiles of 30 full gene knockouts across 14 relevant cell types during human pancreatic development and beta-cell differentiation. pertTF outperforms current methods in predicting expression changes of perturbing unseen genes in unseen cellular contexts. In addition, pertTF infers perturbation-induced shifts in cell identity and population composition, an important phenotypic outcome of perturbation in many physiology and disease settings. Through transfer learning, pertTF operates in physiologically relevant systems, including primary human islets, where large-scale perturbation experiments are challenging. The generalizability of pertTF is further demonstrated by in silico pooled and single-cell CRISPR screens, capturing critical regulators of stem cells and early pancreatic cell development. These results establish pertTF as a framework for integrating large-scale single-cell perturbation data with AI models to predict genetic perturbation effects across cellular systems and disease contexts.

bioinformatics↗

bioRxiv2022

Dynamic states of cervical epithelia during pregnancy and epithelial barrier disruption

The cervical epithelium undergoes continuous changes in proliferation, differentiation, and function that are critical before pregnancy to ensure fertility and during pregnancy to provide a physical and immunoprotective barrier for pregnancy maintenance. Barrier disruption can lead to the ascension of pathogens that elicit inflammatory responses and preterm birth. Here, we identify cervical epithelial subtypes in nonpregnant, pregnant, and in-labor mice using single-cell transcriptome and spatial analysis. We identify heterogeneous subpopulations of epithelia displaying spatial and temporal specificity. Notably, two goblet cell subtypes with distinct transcriptional programs and mucosal networks were dominant in pregnancy. Untimely basal cell proliferation and goblet cells with diminished mucosal integrity characterize barrier dysfunction in mice lacking hyaluronan. These data demonstrate how the cervical epithelium undergoes continuous remodeling to maintain dynamic states of homeostasis in pregnancy and labor, and provide a framework to understand perturbations in epithelial health and host-microbe interactions that increase the risk of premature birth.

cell biology↗

bioRxiv2021

Design considerations of a wearable electronic-skin for mental health and wellness: balancing biosignals and human factors

Chronic stress has been associated with a variety of pathophysiological risks including developing mental illness. Conversely, appropriate stress management, can be used to foster mental wellness proactively. Yet, there is no existing method that accurately and objectively monitors stress. With recent advances in electronic-skin (e-skin) and wearable technologies, it is possible to design devices that continuously measure physiological parameters linked to chronic stress and other mental health and wellness conditions. However, the design approach should be different from conventional wearables due to considerations like signal-to-noise ratio and the risk of stigmatization. Here, we present a multi-part study that combines user-centered design with engineering-centered data collection to inform future design efforts. To assess human factors, we conducted an n=24 participant design probe study that examined perceptions of an e-skin for mental health and wellness as well as preferred wear locations. We complement this with an n=10 and n=16 participant data collection study to measure physiological signals at several potential wear locations. By balancing human factors and biosignals, we conclude that the upper arm and forearm are optimal wear locations.

bioengineering↗

bioRxiv2020

Characterization of interstitial heterogeneity in the developing kidney

Kidney formation requires the coordinated growth of multiple cell types including the collecting ducts, nephrons, vasculature and interstitium. There has been a long-held belief that interactions between the progenitors of the collecting ducts and nephrons are primarily responsible for development of this organ. However, over the last several years, it has become increasingly clear that multiple aspects of kidney development require signaling from the interstitium. How the interstitium orchestrates these various roles is still poorly understood. We show that during development, the interstitium is a highly heterogeneous, patterned population of cells that occupies distinct positions correlated to the adjacent parenchyma. Our analysis indicates that the heterogeneity is not a mere reflection of different stages in a linear developmental trajectory but instead represents several novel differentiated cell states. Further, we find that beta-catenin has a cell autonomous role in the development of a medullary subset of the interstitium and that this non-autonomously affects the development of the adjacent epithelia. These findings suggest the intriguing possibility that the different interstitial subtypes may create microenvironments that play unique roles in development of the adjacent epithelia and endothelia. Graphical AbstractThe developing interstitium is a highly heterogeneous, patterned population of cells that occupies distinct positions correlated to the adjacent parenchyma. O_FIG O_LINKSMALLFIG WIDTH=161 HEIGHT=200 SRC="FIGDIR/small/973966v1_ufig1.gif" ALT="Figure 1"> View larger version (57K): org.highwire.dtl.DTLVardef@1c79dc3org.highwire.dtl.DTLVardef@469934org.highwire.dtl.DTLVardef@f43787org.highwire.dtl.DTLVardef@1698bd6_HPS_FORMAT_FIGEXP M_FIG C_FIG

developmental biology↗

bioRxiv2018

Rational reprogramming of cellular states by combinatorial perturbation

Ectopic expression of transcription factors (TFs) can reprogram cell state. However, due to the large combinatorial space of possible TF cocktails, it remains difficult to identify TFs that reprogram specific cell types. Here, we develop Reprogram-Seq to experimentally screen thousands of TF cocktails for reprogramming performance. Reprogram-Seq leverages organ-specific cell atlas data with single-cell perturbation and computational analysis to predict, evaluate, and optimize TF combinations that reprogram a cell type of interest. Focusing on the cardiac system, we perform Reprogram-Seq on MEFs using an undirected library of 48 cardiac factors and separately on a directed library of 10 epicardial-related TFs. We identify a novel combination of 3 TFs which efficiently reprogram MEFs to epicardial-like cells that are transcriptionally, molecularly, morphologically, and functionally similar to primary epicardial cells. Reprogram-Seq holds promise to accelerate the generation of specific cell types for regenerative medicine.

genomics↗

bioRxiv2018

Frequent sgRNA-barcode Recombination in Single-cell Perturbation Assays

Simultaneously detecting CRISPR-based perturbations and induced transcriptional changes in the same cell is a powerful approach to unraveling genome function. Several lentiviral approaches have been developed, some of which rely on the detection of distally located genetic barcodes as an indirect proxy of sgRNA identity. Since barcodes are often several kilobases from their corresponding sgRNAs, viral recombination-mediated swapping of barcodes and sgRNAs is feasible. Using a self-circularization-based sgRNA-barcode library preparation protocol, we estimate the recombination rate to be ~50% and we trace this phenomenon to the pooled viral packaging step. Recombination is random, and decreases the signal-to-noise ratio of the assay. Our results suggest that alternative approaches can increase the throughput and sensitivity of single-cell perturbation assays.

genomics↗