Biology subjects

Leung, A. K.-Y.

Publications and source records attributed to Leung, A. K.-Y..

8 recordsLinked to original sources

bioRxiv2026

Cooperative and competitive interactions among transcription regulatory elements modulate transcription output

Transcription is modulated by interactions among transcription regulatory elements (TREs), including promoters and enhancers, but their underlying interaction specificities remain opaque. Here, we develop a Chromatin-Integrated, landing-pad-based Enhancer Reporter Assay, CIERA-seq, that interrogates regulatory interactions between target promoters and distal TREs in a controlled genomic and chromatin context. We find both promoter- and enhancer-TREs exhibit enhancer activities on target promoters with a specificity that depends on factors that TREs recruit. Promoters are generally enriched for factors involved in the core transcription process, whereas enhancers are enriched for pioneer factors and chromatin remodelers. Our CIERA-seq assays support a model where TREs activate target promoters constrained at different rate limiting steps in transcription by supplying complementary components of the transcription regulatory machinery. An orthogonal analysis using CRISPRi datasets shows that both promoters and enhancers can positively regulate expression of neighboring genes at their native loci. However, promoters are also more prone than enhancers to exhibit negative regulatory effects, a difference that may arise from competition for limiting transcriptional machinery. Together, these findings support a framework in which promoters and enhancers act as cooperative, or competitive, regulatory elements within local transcription hubs to modulate transcription outputs.

molecular biology↗

bioRxiv2025

High-resolution reconstruction of cell-type specific transcriptional regulatory processes from bulk sequencing samples

Biological systems exhibit remarkable heterogeneity, characterized by intricate interplay among diverse cell types. Resolving the regulatory processes of specific cell types is crucial for delineating developmental mechanisms and disease etiologies. While single-cell sequencing methods such as scRNA-seq and scATAC-seq have revolutionized our understanding of individual cellular functions, adapting bulk genome-wide assays to achieve single-cell resolution of other genomic features remains a significant technical challenge. Here, we introduce Deep-learning-based DEconvolution of Tissue profiles with Accurate Interpretation of Locus-specific Signals (DeepDETAILS), a novel quasi-supervised framework to reconstruct cell-type-specific genomic signals with base-pair precision. DeepDETAILS core innovation lies in its ability to perform cross-modality deconvolution using scATAC-seq reference libraries for other bulk datasets, benefiting from the affordability and availability of scATAC-seq data. DeepDETAILS enables high-resolution mapping of genomic signals across diverse cell types, with great versatility for various omics datasets, including nascent transcript sequencing (such as PRO-cap and PRO-seq) and ChIP-seq for chromatin modifications. Our results demonstrate that DeepDETAILS significantly outperformed traditional statistical deconvolution methods. Using DeepDETAILS, we developed a comprehensive compendium of high-resolution nascent transcription and histone modification signals across 39 diverse human tissues and 86 distinct cell types. Furthermore, we applied our compendium to fine-map risk variants associated with Primary Sclerosing Cholangitis (PSC), a progressive cholestatic liver disorder, and revealed a potential etiology of the disease. Our tool and compendium provide invaluable insights into cellular complexity, opening new avenues for studying biological processes in various contexts.

bioinformatics↗

bioRxiv2025

Comprehensive Evaluation of Diverse Massively Parallel Reporter Assays to Functionally Characterize Human Enhancers Genome-wide

Massively parallel reporter assays (MPRAs) and self-transcribing active regulatory region sequencing (STARR-seq) have revolutionized enhancer characterization by enabling high-throughput functional assessment of regulatory sequences. Here, we systematically evaluated six MPRA and STARR-seq datasets generated in the human K562 cell line and found substantial inconsistencies in enhancer calls from different labs that are primarily due to technical variations in data processing and experimental workflows. To address these variations, we implemented a uniform enhancer call pipeline, which significantly improved cross-assay agreement. While increasing sequence overlap thresholds enhanced concordance in STARR-seq assays, cross-assay consistency in LentiMPRA was strongly influenced by assay-specific factors. Notably, our results show that LentiMPRA exhibits a strong preference for promoter-associated sequences rather than enhancers. Functional validation using candidate cis-regulatory elements (cCREs) confirmed that epigenomic features such as chromatin accessibility and histone modifications are strong predictors of enhancer activity. Importantly, our study validated transcription as a critical hallmark of active enhancers, demonstrating that highly transcribed regions exhibit significantly higher active rates across assays. Furthermore, we show that transcription enhances the predictive power of epigenomic features, enabling more accurate and refined enhancer annotation. Our study provides a comprehensive framework for integrating different enhancer datasets and underscores the importance of accounting for assay-specific biases when interpreting enhancer activity. These findings refine enhancer identification using massively parallel reporter assays and improve the functional annotation of the human genome.

genomics↗

bioRxiv2025

Simultaneous measurement of intrinsic promoter and enhancer potential reveals principles of functional duality and regulatory reciprocity

Accumulating evidence indicates that both promoters and enhancers are capable of exerting promoter and enhancer functions; however, the relationship between these two activities within individual elements and the determinants of this dual functionality remain poorly understood. We developed a massively parallel dual reporter assay that enables simultaneous assessment of the intrinsic promoter and enhancer potential exerted by the same sequence. Parallel quantification for thousands of elements reveals that canonical human promoters and enhancers can act as both promoters and enhancers under the same contexts, and that promoter activity may be necessary but not sufficient for enhancer function. Perturbations to element transcription factor binding motifs lead to disruptions in both activities, implicating a shared syntax for the two regulatory functions. Combinations of elements with different minimal promoters reveal reciprocal activity modulation, which, together with a strong correlation between promoter and enhancer functions, imply a bidirectional feedback loop to sustain environments of high transcriptional activity. Finally, we validate this reciprocity and correlation in situ using CRISPR activation at the human {beta}-globin locus. Our approach reveals that the functional convergence between promoters and enhancers arises from a shared regulatory logic and sequence syntax, advancing a unified model for regulatory element biology.

genomics↗

bioRxiv2023

Accurate identification of structural variations from cancer samples

Structural variations (SVs) are commonly found in cancer genomes. They can cause gene amplification, deletion, and fusion, among other functional consequences. With an average read length of hundreds of kilobases, nano-channel-based optical DNA mapping is powerful in detecting large SVs. However, existing SV calling methods are not tailored for cancer samples, which have special properties such as mixed cell types and sub-clones. Here we propose the COMSV method that is specifically designed for cancer samples. It shows high sensitivity and specificity in benchmark comparisons. Applying to cancer cell lines and patient samples, COMSV identifies hundreds of novel SVs per sample.

genomics↗

bioRxiv2021

Systematic comparison of experimental assays and analytical pipelines for identification of active enhancers genome-wide

Mounting evidence supports the idea that transcriptional patterns serve as more specific identifiers of active enhancers than histone marks1,2; however, the optimal strategy to identify active enhancers both experimentally and computationally has not been determined. In this study, we compared 13 genome-wide RNA sequencing assays in K562 cells and showed that the nuclear run-on followed by cap-selection assay (namely, GRO/PRO-cap) has significant advantages in eRNA detection and active enhancer identification. We also introduced a new analytical tool, Peak Identifier for Nascent-Transcript Sequencing (PINTS), to identify active promoters and enhancers genome-wide and pinpoint the precise location of the 5 transcription start sites (TSSs) within these regulatory elements. Finally, we compiled a comprehensive enhancer candidate compendium based on the detected eRNA TSSs available in 120 cell and tissue types. To facilitate the exploration and prioritization of these enhancer candidates, we also built a user-friendly web server (https://pints.yulab.org) for the compendium with various additional genomic and epigenomic annotations. With the knowledge of the best available assays and pipelines, this large-scale annotation of candidate enhancers will pave the road for selection and characterization of their functions in a time-, labor-, and cost-effective manner in the future.

genomics↗

bioRxiv2019

MaXLinker: proteome-wide cross-link identifications with high specificity and sensitivity

Protein-protein interactions play a vital role in nearly all cellular functions. Hence, understanding their interaction patterns and three-dimensional structural conformations can provide crucial insights about various biological processes and underlying molecular mechanisms for many disease phenotypes. Cross-linking mass spectrometry has the unique capability to detect protein-protein interactions at a large scale along with spatial constraints between interaction partners. However, the current cross-link search algorithms follow an "MS2-centric" approach and, as a result, suffer from a high rate of mis-identified cross-links (~15%). We address this urgent problem, by designing a novel "MS3-centric" approach for cross-link identification and implemented it as a search engine called MaXLinker. MaXLinker significantly outperforms the current state of the art search engine with up to 18-fold lower false positive rate. Additionally, MaXLinker results in up to 31% more cross-links, demonstrating its superior sensitivity and specificity. Moreover, we performed proteome-wide cross-linking mass spectrometry using K562 cells. Employing MaXLinker, we unveiled the most comprehensive set of 9,319 unique cross-links at 1% false discovery rate, comprising 8,051 intraprotein and 1,268 interprotein cross-links. Finally, we experimentally validated the quality of a large number of novel interactions identified in our study, providing a conclusive evidence for MaXLinkers robust performance.

systems biology↗

bioRxiv2017

OMSV enables accurate and comprehensive identification of large structural variations from nanochannel-based single-molecule optical maps

Human genomes contain structural variations (SVs) that are associated with various phenotypic variations and diseases. SV detection by sequencing is incomplete due to limited read length. Nanochannel-based optical mapping (OM) allows direct observation of SVs up to hundreds of kilo-bases in size on individual DNA molecules, making it a promising alternative technology for identifying large SVs. SV detection from optical maps is non-trivial due to complex types of error present in OM data, and no existing methods can simultaneously handle all these complex errors and the wide spectrum of SV types. Here we present a novel method, OMSV, for accurate and comprehensive identification of SVs from optical maps. OMSV detects both homozygous and heterozygous SVs, SVs of various types and sizes, and SVs with and without creating/destroying restriction sites. In an extensive series of tests based on real and simulated data, OMSV achieved both high sensitivity and specificity, with clear performance gains over the latest existing method. Applying OMSV to a human cell line, we identified hundreds of SVs >2kbp, with 65% of them missed by sequencing-based callers. Independent experimental validations confirmed the high accuracy of these SVs. We also demonstrate how OMSV can incorporate sequencing data to determine precise SV break points and novel sequences in the SVs not contained in the reference. We provide OMSV as open-source software to facilitate systematic studies of large SVs.

bioinformatics↗