Biology subjects

Schiefelbein, J.

Publications and source records attributed to Schiefelbein, J..

8 recordsLinked to original sources

bioRxiv2025

scCoBench: Benchmarking single cell RNA-seq co-expression using promoter-reporter lines

Single-cell RNA sequencing (scRNA-seq) has become a powerful tool for uncovering transcriptomic heterogeneity and reconstructing gene regulatory networks in complex tissues. However, the sparsity, high noise levels, and dropout events inherent to scRNA-seq data pose challenges for accurate inference of gene-gene relationships. In this study scCoBench, we systematically benchmark correlation metrics, pseudo bulk analysis, and imputation methods using promoter-reporter and native gene pairs as internal controls to evaluate the performance of ten widely used gene-gene co-expression measurements. Interestingly, we found that commonly used data scaling and normalization approaches lead to lower correlation between promoter reporter and native gene pairs in most of the co-expression methods. Moreover, we assess the impact of five popular imputation techniques, including scImpute, SAVER, Autoencoder (AE), Variational Autoencoder (VAE), and Generative Adversarial Network (GAN) on recovering biologically relevant co-expression patterns. Our results demonstrate that imputation models not only markedly enhance correlation between each promoter-reporter and native gene pair but also increase the number of cells co-expressing both genes. Imputation also improved transcription factor target gene correlations and revealed stronger associations among genes within the same protein complex. This work highlights the utility of promoter-reporter systems for benchmarking computational methods and underscores the potential of deep learning-based imputation to improve the biologically relevant signals of scRNA-seq data.

bioinformatics↗

bioRxiv2024

Diversification of gene expression across extremophytes and stress-sensitive species in the Brassicaceae

Plants are composed of diverse cell types that facilitate adaptations to the environment, yet cross-species comparisons of such response programs at single-cell resolution remain scarce. To explore this diversity, here we profiled >200,000 root cells from five Brassicaceae species, including stress-sensitive species (Arabidopsis thaliana, Sisymbrium irio), extremophytes (Eutrema salsugineum, Schrenkiella parvula), and the polyploid crop Camelina sativa under control, NaCl, and abscisic acid (ABA) treatments. We developed a computational pipeline to characterize the conservation and divergence of cell-type gene expression across the Brassicaceae, revealing that approximately half of previously defined Arabidopsis cell-type markers fail to maintain conserved expression in one or more non-Arabidopsis species. Therefore, we curated a refined set of pan-Brassicaceae markers and identified orthologs whose expression profiles have diverged across lineages. Using in situ hybridization, we mapped distinct cortex subpopulations to specific cortical layers across species, and found flavonoid biosynthesis programs preferentially localized to the inner cortex layer, linking cortex-layer specification to metabolic specialization. Cell-type contributions to stress responses differed among species and across treatments, with lineage-specific losses of responsiveness occurring less frequently but evolutionarily more favored than lineage-specific gains. In C. sativa, sub-genomes contributed equally to stress responses, and homeologs with divergent responses typically lacked other signatures of functional divergence. Together, this work establishes a foundational root single-cell atlas and an analytical framework for multi-species comparative transcriptomics, providing insights into how stress responses diversify across cell types, stress-sensitive to stress-adapted species, and in crop lineages.

plant biology↗

bioRxiv2023

TTG2 defines trichome cell shape by modulating microtubules and actin cytoskeleton in Arabidopsis

The Arabidopsis TRANSPARENT TESTA GLABRA2 (TTG2) gene encodes a WRKY transcription factor that regulates a range of development events like trichome, seed coat, and atrichoblasts formation. Loss-of-function of TTG2 was previously shown to reduce or eliminate trichome branching. Here, we report the identification of a new allele of TTG2, ttg2-6. In contrast to the ttg2 mutants described before, ttg2-6 displays unique trichome phenotypes. Some ttg2-6 mutant trichomes are hyperbranched, whereas others are hypobranched, distorted, or clustered. Further, we found that in addition to specifically activating R3 MYB transcription factor TRIPTYCHON (TRY) to promote trichome branching, TTG2 also integrates cytoskeletal signaling to regulate trichome morphogenesis. The ttg2-6 trichomes display aberrant cortical microtubules (cMTs) and cortical actin filaments (F-actin) configurations. Moreover, genetic and biochemical analysis showed that TTG2 could directly bind and regulate the expression of BRICK1 (BRK1), a subunit of the actin nucleation promoting complex SCAR/WAVE. Collectively, taking advantage of ttg2-6, we uncovered new functions for TTG2 in facilitating cMT and F-actin cytoskeleton-dependent trichome development. Our evidence for a direct relationship between TTG2 and cytoskeletal regulators establishes an unprecedented understanding of cellular signaling events downstream of the core transcriptional regulation during trichome development in Arabidopsis.

developmental biology↗

bioRxiv2023

A conserved gene regulatory network controls root epidermal cell patterning in superrosid species

O_LIIn superrosid species, root epidermal cells differentiate into root hair cells and non-hair cells. In some superrosids, the root hair cells and non-hair cells are distributed randomly (Type I pattern) and in others, they are arranged in a position-dependent manner (Type III pattern). The model plant Arabidopsis (Arabidopsis thaliana) adopts the Type III pattern, and the gene regulatory network (GRN) that controls this pattern has been defined. However, it is unclear whether the Type III pattern in other species is controlled by a similar GRN as in Arabidopsis, and it is not known how the different patterns evolved. C_LIO_LIIn this study, we analyzed superrosid species Rhodiola rosea, Boehmeria nivea, and Cucumis sativus for their root epidermal cell patterns. Combining phylogenetics, transcriptomics, and cross-species complementation, we analyzed homologs of the Arabidopsis patterning genes from these species. C_LIO_LIWe identified R.rosea and B.nivea as Type III species and C.sativus as Type I species. We discovered substantial similarities in structure, expression, and function of Arabidopsis patterning gene homologs in R.rosea and B.nivea, and major changes in C.sativus. C_LIO_LIWe propose that in superrosids, diverse Type III species inherited the patterning GRN from a common ancestor, whereas Type I species arose by mutations in multiple lineages. C_LI

plant biology↗

bioRxiv2020

Identification of cell-type marker genes from plant single-cell RNA-seq data using machine learning

An essential step in the analysis of single-cell RNA sequencing data is to classify specific cell types with marker genes. In this study, we have developed a machine learning pipeline called Single cell Predictive markers (SPmarker) to identify novel cell-type marker genes in the Arabidopsis root. Unlike traditional approaches, our method uses interpretable machine learning methods to select marker genes. We have demonstrated that our method can (1) assign cell types based on cells that were labeled using published methods, (2) project cell types identified by trajectory analysis from one dataset to other datasets, and (3) assign cell types based on internal GFP markers. Using SPmarker, we have identified hundreds of new marker genes that were not identified before. As compared to known marker genes, we have found more orthologous genes of these new marker genes in corresponding rice single cell clusters. We have also found 172 new marker genes for trichoblast in five non-Arabidopsis species, which expands number of marker genes for this cell type by 35-154%. Our results represent a new approach to identify cell-type marker genes from scRNA-seq data and pave the way for cross-species mapping of scRNA-seq data in plants.

bioinformatics↗

bioRxiv2020

Expression partitioning of duplicate genes at single cell resolution in Arabidopsis roots

Gene duplication is a key evolutionary phenomenon, prevalent in all organisms but particularly so in plants, where whole genome duplication (WGD; polyploidy) is a major force in genome evolution. Much effort has been expended in attempting to understand the evolution of duplicate genes, addressing such questions as why some paralogue pairs rapidly return to single copy status whereas, in other pairs, paralogues are retained and may (or may not) diverge in expression pattern or function. The effect of a gene--its site of expression and thus the initial locus of its function--occurs at the level of a cell comprising a single cell type at a given state of the cells development. Thus, it is critical to understand the expression of duplicated gene pairs at a cellular level of resolution. Using Arabidopsis thaliana root single cell transcriptomic data we identify 36 cell clusters, each representing a cell type at a particular developmental state, and analyze expression patterns of over 11,000 duplicate gene pairs produced by three cycles of polyploidy as well as by various types of single gene duplication mechanisms. We categorize paralogue pairs by their patterns of expression, identifying pairs showing strongly biased paralogue/homoeologue expression in different cell clusters. Notably, the precision of cell-level expression data permits the identification of pairs showing alternate bias, with each paralogue comprising 90% or greater of the pairs expression in different cell clusters, consistent with subfunctionalization at the cell type or cell state level, and, in some cases, at the level of individual cells. We identify a set of over 7,000 genes whose expression in all 36 cell clusters suggests that the single copy ancestor of each was also expressed in all root cells. With this cell-level expression information we hypothesize that there have been major shifts in expression for the majority of duplicated genes, to different degrees depending, as expected, on gene function and duplication type, but also on the particular cell type and state.

plant biology↗

bioRxiv2020

The impact of chromatin remodeling on gene expression at the single cell level in Arabidopsis thaliana

Similar to other complex organisms, plants consist of diverse and highly specialized cell types. The gain of unique biological functions of these different cell types is the consequence of the establishment of cell-type-specific transcriptional programs and their associated regulatory mechanisms. Recently, single cell transcriptomic approaches have been applied on Arabidopsis thaliana root protoplasts allowing the accurate characterization of the transcriptional profiles of the cell-types composing seedling roots. As a first step in gaining a deeper understanding of the regulatory mechanisms controlling Arabidopsis gene expression, we report the use of single nucleus RNA sequencing (sNucRNA-seq) and single nucleus Assay for Transposase Accessible Chromatin sequencing (sNucATAC-seq) technologies on Arabidopsis roots. The comparison of our single nuclei transcriptomes to previously published protoplast transcriptomes validated the use of nuclei as biological entities to establish cell-type specific transcriptomes from multicellular organs. Furthermore, our sNucRNA-seq results uncovered the transcriptome of additional cell subtypes not identified by scRNA-seq. Similar to our transcriptomic approach, the sNucATAC-seq approach led to the distribution of the Arabidopsis nuclei into distinct clusters suggesting the differential remodeling of the chromatin between groups of cells according to their identity. To reveal the impact of chromatin remodeling on gene transcription, we integrated sNucRNA-seq and sNucATAC-seq data and demonstrated that cell-type-specific marker genes also display cell-type-specific pattern of chromatin accessibility. Our data suggest that the differential remodeling of the chromatin is a critical mechanism to regulate gene activity at the cell-type level.

plant biology↗

bioRxiv2019

Self-supervised retinal thickness prediction enables deep learning from unlabeled data to boost classification of diabetic retinopathy

Access to large, annotated samples represents a considerable challenge for training accurate deep-learning models in medical imaging. While current leading-edge transfer learning from pre-trained models can help with cases lacking data, it limits design choices, and generally results in the use of unnecessarily large models. We propose a novel, self-supervised training scheme for obtaining high-quality, pre-trained networks from unlabeled, cross-modal medical imaging data, which will allow for creating accurate and efficient models. We demonstrate this by accurately predicting optical coherence tomography (OCT)-based retinal thickness measurements from simple infrared (IR) fundus images. Subsequently, learned representations outperformed advanced classifiers on a separate diabetic retinopathy classification task in a scenario of scarce training data. Our cross-modal, three-staged scheme effectively replaced 26,343 diabetic retinopathy annotations with 1,009 semantic segmentations on OCT and reached the same classification accuracy using only 25% of fundus images, without any drawbacks, since OCT is not required for predictions. We expect this concept will also apply to other multimodal clinical data-imaging, health records, and genomics data, and be applicable to corresponding sample-starved learning problems.

bioinformatics↗