Biology subjects

Bergmann, T.

Publications and source records attributed to Bergmann, T..

5 recordsLinked to original sources

bioRxiv2026

Unsupervised Clustering Enables Cohort-Specific Variant Recalibration in Crops

Genomic selection (GS) and trait discovery in crops require reliable marker datasets, yet variant discovery remains constrained by the lack of standardised, high-confidence reference resources. Publicly available plant genomic resources are often based on outdated reference genomes, heterogenous pipelines, and generic variant hard-filtering thresholds derived from human studies. Consequently, implementation of the Genome Analysis Toolkit (GATK) Best Practices framework, including machine learning-based variant quality score recalibration (VQSR), remains challenging in crop genomes. Here, we present a crop-adapted framework for unsupervised variant quality assessment and VQSR enabling advanced SNP and insertion-deletion (INDEL) discovery without predefined high-confidence reference variant sets. Using Brassica napus as a model, we use the Illumina 60k Brassica SNP array to generate a reference-specific truth set for data pre-processing and independent evaluation of an unsupervised machine-learning approach for constructing cohort-specific VQSR resources. Gaussian mixture model (GMM)-based clustering of multidimensional variant annotation profiles identifies high-, intermediate-, and low-quality variant classes, providing flexible training resources for VQSR without requiring labelled training data. We further validate the framework in diploid Brassica oleracea and hexaploid wheat (Triticum aestivum) across cohorts differing in population size, genetic diversity, and genome architecture. Despite population- and species-specific differences in variant annotation profiles, GMM-based clustering consistently identifies informative variant quality classes and enables cohort-specific recalibration of both SNPs and INDELs. Our results demonstrate that unsupervised machine learning can extend GATK Best Practices to crop species lacking comprehensive genomic resources, providing a scalable approach for generating high-quality variant datasets for crop genomics and downstream breeding applications.

plant biology↗

bioRxiv2026

Capturing the missing gene content of elite wheat by analysis of wheat landraces

There is a need to breed superior wheat cultivars to meet increasing global food demand. However, modern cultivars have undergone a substantial loss of genetic diversity due to intensive breeding. A pool of genetic diversity remains untapped in wheat landraces, and pangenomics can help identify genes of potential agronomic importance in these old lines that can be applied to accelerate wheat improvement. Here, we have constructed the largest wheat pangenome to date, representing 1,061 diverse individuals (827 landraces and 234 modern cultivars) from 47 countries across six continents. We identified 10,426 predicted gene models specific to landraces that are enriched for functions associated with disease resistance, abiotic stress adaptation, and symbiosis. Our results reveal that landraces harbour an extensive repertoire of genes that are absent from modern wheat cultivars and provide a foundational resource for the systematic reintroduction of adaptive variation to enhance wheat resilience and sustainability in the face of climate change.

plant biology↗

bioRxiv2023

PenTag, a Versatile Platform for Covalent Bioconjugation, Purification, and Tagging of Proteins Towards the Development of Novel Biohybrid Material Systems

The site-specific and covalent conjugation of proteins on solid supports and in hydrogels is the basis for the synthesis of biohybrid materials offering broad applications. Current methods for conjugating proteins to desired targets are often challenging due to unspecific binding, unstable (non-covalent) coupling, or expensive and difficult-to-synthesize ligand molecules. Here, we present PenTag, an approach for the biorthogonal, highly-specific and covalent conjugation of a protein to its ligand for various applications in materials sciences. We engineered penicillin- binding protein 3 (PBP3) and showed that this protein can be used for the stable and spontaneous conjugation of proteins to dyes, polymers, or solid supports. We applied PenTag as a crosslinking tool for synthesizing stimuli-responsive hydrogels or for the development of a biohybrid material system performing computational operations emulating a 4:2 encoder. Based on this broad applicability and the use of a small, cheap and easy-to-functionalize ligand and a stable, soluble recombinant protein, we see PenTag as a versatile approach towards biohybrid material synthesis.

synthetic biology↗

bioRxiv2021

Development of the entorhinal cortex occurs via parallel lamination during neurogenesis.

The entorhinal cortex (EC) is the spatial processing center of the brain and structurally is an interface between the three layered paleocortex and six layered neocortex, known as the periarchicortex. Limited studies indicate peculiarities in the formation of the EC such as early emergence of cells in layers (L) II and late deposition of LIII, as well as divergence in the timing of maturation of cell types in the superficial layers. In this study, we examine developmental events in the entorhinal cortex using an understudied model in neuroanatomy and development, the pig and supplement the research with BrdU labeling in the developing mouse EC. We determine the pig serves as an excellent anatomical model for studying human neurogenesis, given its long gestational length, presence of a moderate sized outer subventricular zone and early cessation of neurogenesis during gestation. Immunohistochemistry identified prominent clusters of OLIG2+ oligoprogenitor-like cells in the superficial layers of the lateral EC (LEC) that are sparser in the medial EC (MEC). These are first detected in the subplate during the early second semester. MRI analyses reveal an acceleration of EC growth at the end of the second trimester. BrdU labeling of the developing MEC, shows the deeper layers form first and prior to the superficial layers, but the LV/VI emerges in parallel and the LII/III emerges later, but also in parallel. We coin this lamination pattern parallel lamination. The early-born Reln+ stellate cells in the superficial layers express the classic LV marker, Bcl11b (Ctip2) and arise from a common progenitor that forms the late deep layer LV neurons. In summary, we characterize the developing EC in a novel animal model and outline in detail the formation of the EC. We further provide insight into how the periarchicortex forms in the brain, which differs remarkably to the inside-out lamination of the neocortex.

developmental biology↗

bioRxiv2019

Sandwich cortical lamination and single-cell analysis decodes the developing spatial processing system.

Classic studies investigating how and when the entorhinal cortex (component of the memory processing system of the brain) develops have been based on traditional thymidine autoradiography and histological techniques. In this study, we take advantage of modern technologies to trace at a high resolution, the cellular complexity of the developing porcine medial entorhinal cortex by using single-cell profiling. The postnatal medial entorhinal cortex comprises 4 interneuron, 3 pyramidal neuron and 2 stellate cell populations which emerge from intermediate progenitor and immature neuron populations. We discover four MGE-derived interneurons and one CGE-derived interneuron population as well as several IN progenitors. We also identify two oligodendrocyte progenitor populations and three populations of oligodendrocytes. We perform a proof-of-concept experiment demonstrating that porcine scRNA-seq data can be used to develop novel protocols for producing human entorhinal cells in-vitro. We identified six transcription factors (RUNX1A1, SOX5, FOXP1, MEF2C, TCF4, EYA2) important in neurodevelopment and differentiation from one RELN+ stellate cell population. Using a lentiviral vector approach, we reprogrammed human induced pluripotent stem cells into stellate cell-like cells which expressed RELN, SATB2, LEF1 and BCL11B. Our findings contribute to the understanding of the formation of the brains cognitive memory and spatial processing system and provides proof-of-concept for the production of entorhinal cells from human pluripotent stem cells in-vitro.

developmental biology↗