Biology subjects

Radley, A.

Publications and source records attributed to Radley, A..

7 recordsLinked to original sources

bioRxiv2026

ESFS: A Noise-Resilient Framework for Feature Selection and Marker Gene Discovery in Single-Cell Transcriptomics

Single-cell RNA sequencing (scRNA-seq) has transformed our ability to resolve cellular heterogeneity, but extracting meaningful signals remains challenging due to technical noise and batch effects. Most methods for denoising scRNA-seq data have focused on using latent representations such as principal component analysis and deep learning to prioritise biological signals. By contrast, despite its influence on downstream analyses, feature selection has received relatively limited attention, leading to widespread reliance on the comparatively simplistic strategy of highly variable gene selection. Here we present Entropy Sorting Feature Selection (ESFS), a modular, user-friendly framework that substantially improves the interpretability of scRNA-seq data. Notably, ESFS reveals complex expression dynamics that are obscured in latent representations. We demonstrate the utility of ESFS in diverse data: identifying coherent developmental programs across eight independent human embryo datasets without batch integration; resolving spatial gene expression in mouse colon missed by conventional analyses; disambiguating shared and tumour-specific microenvironments in glioblastoma; and disentangling spatial, temporal, and neurogenic programs in the developing mouse neural tube. Beyond delivering a powerful and user-friendly software that deepens insight into complex biological systems, our work establishes Entropy Sorting as a novel information theoretic for advanced data analysis methods.

bioinformatics↗

bioRxiv2025

Hierarchical lineage architecture of human and avian spinal cord revealed by single-cell genomic barcoding

The formation of neural circuits depends on the precise spatial and temporal organisation of neuronal populations during development. In the vertebrate spinal cord, progenitors are patterned into molecularly defined domains, but how lineage relationships shape neuronal diversity and function has remained unclear. Here, we combine genomic barcoding with single- cell RNA sequencing in chick and human embryos to generate cell-type-resolved clonal maps. We find that spinal neurogenesis follows a hierarchical organisation in which the neural tube first partitions into five broad subdivisions that resolve into the eleven progenitor domains generating the cardinal neuronal classes. This bifurcating architecture implies a patterning mechanism of sequential binary decisions. The most prominent lineage restriction occurs at the embryonic alar-basal boundary, separating sensory-processing from motor-control circuits. Individual progenitors generate neurons across multiple temporal waves while remaining constrained within their lineage subdivision, demonstrating persistence of spatial identity despite temporal competence changes. Among sensory populations, we identify two developmental routes, via unifated or bifated progenitors, to pain- and itch-processing interneurons. These principles are conserved between chick and human, with clonal analysis in human embryos revealing that most fate choices are resolved by six weeks post-conception. Together, these findings provide a framework for spinal cord development and reveal lineage compartmentalisation as a fundamental principle in neural circuit assembly and evolution.

developmental biology↗

bioRxiv2024

An in vivo CRISPR screen in chick embryos reveals a role for MLLT3 in specification of neural cells from the caudal epiblast

Tissue development relies on the coordinated differentiation of stem cells in dynamically changing environments. The formation of the vertebrate neural tube from stem cells in the caudal lateral epiblast (CLE) is a well characterized example. Despite an understanding of the signalling pathways involved, the gene regulatory mechanisms remain poorly defined. To address this, we developed a multiplexed in vivo CRISPR screening approach in chick embryos targeting genes expressed in the caudal epiblast and neural tube. This revealed a role for MLLT3, a component of the super elongation complex, in the specification of neural fate. Perturbation of MLLT3 disrupted neural tube morphology and reduced neural fate acquisition. Mutant forms of Retinoic Acid Receptor A lacking the MLLT3 binding domain similarly reduced neural fate acquisition. Together, these findings validate an in vivo CRISPR screen strategy in chick embryos and identify a previously unreported role for MLLT3 in caudal neural tissue specification.

developmental biology↗

bioRxiv2023

Human naive stem cell models reveal the role of FGF signalling in hypoblast specification in the human embryo

The hypoblast is an essential extra-embryonic tissue set aside within the inner cell mass early in mammalian embryo development, in the blastocyst. Research with human embryos is challenging. Thus, stem cell models that reproduce hypoblast differentiation provide valuable alternatives. We show here that human naive PSC to hypoblast differentiation proceeds via reversion to a transitional ICM-like state, from which hypoblast emerges in concordance with the trajectory in human blastocysts. We identified a window when fibroblast growth factor (FGF) signalling is critical for hypoblast specification. Revisiting FGF signalling in human embryos revealed that inhibition in the early blastocyst suppresses hypoblast formation. In vitro, the induction of hypoblast is synergistically enhanced by limiting trophectoderm and epiblast fates. This finding revises previous reports and establishes a conservation in lineage specification between mouse and human. Overall, this study demonstrates the utility of human naive PSC-based models in elucidating mechanistic features of early human embryogenesis.

developmental biology↗

bioRxiv2023

Branching topology of the human embryo transcriptome revealed by entropy sort feature weighting

Single cell transcriptomics (scRNA-seq) transforms our capacity to define cell states and reveal developmental trajectories. Resolution is challenged, however, by high dimensionality and noisy data. Analysis is therefore typically performed after sub-setting to highly variable genes (HVGs). However, existing HVG selection techniques have been found to have poor agreement with one another, and tend to be biased towards highly expressed genes. Entropy sorting provides an alternative mathematical framework for feature subset selection. Here we implement continuous entropy sort feature weighting (cESFW). On synthetic datasets, cESFW outperforms HVG selection in distinguishing cell state specific genes. We apply cESFW to six merged scRNA-seq datasets spanning human early embryo development. Without smoothing or augmenting the raw counts matrices, cESFW generates a high-resolution embedding displaying coherent developmental progression from 8-cell to post-implantation stages, delineating 15 distinct cell states. The embedding highlights sequential lineage decisions during blastocyst development while unsupervised clustering identifies branch point populations. Cells previously claimed to lack a developmental trajectory reside in the first branching region where morula differentiates into Inner Cell Mass (ICM) or Trophectoderm (TE). We quantify the relatedness of pluripotent stem cell cultures to embryo cell types and identify naive and primed marker genes conserved across culture conditions and the human embryo. Finally, by identifying genes with specifically enriched and dynamic expression during blastocyst formation, we provide markers for staging lineage progression from morula to blastocyst. Together these analyses indicate that cESFW provides the ability to reveal gene expression dynamics in scRNA-seq data that HVG selection can fail to elucidate.

bioinformatics↗

bioRxiv2022

Functional feature selection reveals the inner cell mass in human pre-implantation embryo single cell RNA sequencing data

A major challenge in single cell gene expression analysis is to discern meaningful cellular heterogeneity from technical or biological noise. To address this challenge, we present Entropy Sorting, a mathematical framework that distinguishes genes indicative of cell identity. ES achieves this in an unsupervised manner by quantifying if observed correlations between features are more likely to have occurred due to random chance versus a dependent relationship, without the need for any user defined significance threshold. On synthetic data we demonstrate the removal of noisy signals to reveal a higher resolution of gene expression patterns than commonly used feature selection methods. We then apply ES to human pre-implantation embryo scRNA-seq data. Previous studies failed to unambiguously identify early inner cell mass (ICM), suggesting that the human embryo may diverge from the mouse paradigm. In contrast, ES resolves the ICM and reveals sequential lineage bifurcations as in the classical model. Entropy sorting thus provides a powerful approach for maximising information extraction from high dimensional datasets such as scRNA-seq data.

bioinformatics↗

bioRxiv2018

Synthetic chromosome fusion: effects on genome structure and function

As part of the Synthetic Yeast 2.0 (Sc2.0) project, we designed and synthesized synthetic chromosome I. The total length of synI is [~]21.4% shorter than wild-type chromosome I, the smallest chromosome in Saccharomyces cerevisiae. SynI was designed for attachment to another synthetic chromosome due to concerns of potential instability and karyotype imbalance. We used a variation of a previously developed, robust CRISPR-Cas9 method to fuse chromosome I to other chromosome arms of varying length: chrIXR (84kb), chrIIIR (202kb) and chrIVR (1Mb). All fusion chromosome strains grew like wild-type so we decided to attach synI to synIII. Through the investigation of three-dimensional structures of fusion chromosome strains, unexpected loops and twisted structures were formed in chrIII-I and chrIX-III-I fusion chromosomes, which depend on silencing protein Sir3. These results suggest a previously unappreciated 3D interaction between HMR and the adjacent telomere. We used these fusion chromosomes to show that axial element Red1 binding in meiosis is not strictly chromosome size dependent even though Red1 binding is enriched on the three smallest chromosomes in wild-type yeast, and we discovered an unexpected role for centromeres in Red1 binding patterns.

synthetic biology↗