Focus on: 
Back to Conference View
Login

Summer School on Gene Expression across Scales

→ Europe/Berlin
Hahn-Hörsaal (Harnack-Haus)

Hahn-Hörsaal

Harnack-Haus

Ihnestrasse 16-20 - 14195 Berlin, Germany
Description

The IMPRS Summer School on "Gene Expression across Scales" offers a series of research talks by invited speakers, lectures on cutting-edge technologies and uniquely designed workshops that bridge various methodological scales for studying the regulation of gene expression. Topics range from systems biology and big data analysis to mechanistic structural biology and high-resolution imaging.

A key part of this summer school is the interaction between invited speakers and participants. Accordingly, all participants will have to present their research in a poster format, and some selected participants will be invited to give a talk based on their poster abstract.

The summer school is a collaboration between the International Max Planck Research School for Biology and Computation (IMPRS-BAC, MPIMG, Berlin) and the International Max Planck Research School for Genome Science (IMPRS-GS, MPINAT, Göttingen) graduate programs. The number of seats is limited, and priority will be given to doctoral candidates from both IMPRS-BAC and IMPRS-GS.

Confirmed speakers

Fees

There is no registration fee, and meals will be provided. Travel and accommodation costs for IMPRS-GS participants will be covered.

 

Application deadline

Wednesday, 19th of July 2026 (23:59 CEST)

Programme booklet
Contact
  • Monday 7 September
    • 08:30 → 09:00
      Registration 30m Lobby

      Lobby

      Harnack-Haus
      Speakers: Dr Anne-Dominique Gindrat (MPI für molekulare Genetik), Henriette Irmer (MPI-nat)
    • 09:00 → 09:15
      Opening 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany
      Speakers: Dr Anne-Dominique Gindrat (MPI für molekulare Genetik), Henriette Irmer (MPI-nat)
    • 09:15 → 10:30
      Research talk: Mapping the central dogma with multiplexed imaging 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Eva Verzani

      Abstract:
      Gene regulation is a complex process that translates a single genetic code into the multitude of cell types that define an organism. One particularly hard to access layer of this regulation happens at the level of DNA folding non-randomly into the nucleus. The onset of capture-based techniques has highlighted the importance of biophysical interactions of DNA across multiple scales. However, most of these methods have limitations retaining tissue context and measuring beyond simple pairwise interactions. Advances in multiplexed imaging approaches have allowed us to both surpass the diffraction limit of light and the technical limitation of fluorescent imaging channels, enabling multi-scale interrogation of chromatin organization and transcription in native tissue contexts. This talk will focus on tissue-scale imaging of transcriptional state and chromatin organization utilizing Optical Reconstruction of Chromatin Architecture to better connect biophysical properties of chromatin organization to cell state.

      Speaker: Sedona Murphy (Max Planck Institute for Molecular Genetics)
    • 10:30 → 11:00
      Coffee break 30m Lobby

      Lobby

      Harnack-Haus
    • 11:00 → 13:00
      Selected student talks 1-3 2h Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Eva Verzani

      Speakers: Abeera Fatima (Herwig Lab (MPIMG)), Cesar Mateo Bastidas Betancourt (German Primate Center), Siddharth Murali (Georg-August-University Göttingen)
      • CoreNet - coreness-aware node embedding to improve disease module discovery 40m

        Abstract:
        Identifying disease-associated gene modules from omics data is challenging due to data complexity and the difficulty of prioritizing functionally relevant genes, what requires robust network approaches. Such approaches, for example node2vec, use biased random walks to generate node embeddings to map network genes to vector spaces that preserve network neighborhood. However, standard random-walk methods overlook higher-order topological features such as node coreness, which measures centrality within dense network regions. We present a novel workflow that biases random walks using coreness to generate node embeddings via the word2vec skip-gram model. We applied the method to genome-wide association studies (GWAS) and show that our approach produces embeddings that clearly separate trait-associated from non-trait associated genes. By connecting trait-associated genes with influential high-core nodes, we extract functional modules with significantly improved pathway enrichment. This scalable framework enhances precision network biology and has broad applications in gene prioritization and therapeutic target discovery, revealing disease mechanisms and prioritizing therapeutic targets.

        Speaker: Abeera Fatima (Herwig Lab (MPIMG))
      • Genomics of Adaption in European Drosophila melanogaster populations 40m

        Abstract:
        Organisms are adapted to their local conditions and thus, populations differ across geographical distances. Despite major advances in revealing genotype-phenotype associations, the biological processes affected by genomic differences are still largely unknown. Thus, for many traits and model systems a thorough mechanistic understanding of genotype-phenotype associations for complex traits is lacking.

        The European Drosophila Population Genomics Consortium (DrosEU) has recently established a resource of 160 isofemale lines for 9 natural Drosophila melanogaster populations across Europe. The lines had been distributed among multiple laboratories for an unprecedented phenotyping effort wherein 18 organismal traits including life history (e.g. viability, developmental time, fecundity, lifespan), morphology (e.g. thorax length, wing area, wing shape, pigmentation), physiology (e.g. dry weight, cold-shock and heat-shock mortality, chill-coma recovery time) and behavioural traits (circadian eclosion timing, circadian locomotor activity) were quantified. Among these, candidate traits were identified that show signatures for local adaptation.

        To link genomic variants to variation in organismic traits, the genomes of all 160 isofemale lines were sequenced. Population-genetic analyses revealed the existence of geographical structure in the sequence data and identified various candidate loci associated with the measured traits. RNA was extracted at different life stages (by gender) of the same lines, thus adding a layer of gene-expression information to complement GWAS candidates. For the same life-stages, ATAC-seq was also performed to obtain gene-regulatory information of the differentially expressed genes.

        Eventually, the multi-omics data will be integrated to reveal high-confidence loci responsible for local adaptation.

        Speaker: Siddharth Murali (Georg-August-University Göttingen)
      • Comparative primate transcriptomics identifies a ZNF90–OVOL2 regulatory axis shaping human neural progenitor cell dynamics 40m

        Abstract:
        Understanding the behavior of human neural progenitor cells (NPCs) requires comparative characterization of their transcriptomic landscape, particularly through comparisons with closely related primate species. The rhesus macaque represents a key non-hominoid outgroup for such analyses, and several representative transcriptomic datasets are now available. Here, we compare genes enriched in human NPCs with those enriched in rhesus macaque NPCs and found that human NPC-enriched genes are associated with gene programs supporting radial glial identity and proliferative capacity, as well as functions related to sister chromatid segregation. This analysis identified two zinc-finger transcription factors for functional investigation: the ape-specific ZNF90 and the highly conserved OVOL2. By analysing their genome-wide binding, transcriptional output, and cellular effects in cerebral organoids, we uncovered a previously uncharacterized regulatory axis with divergent but partially overlapping effects that converge on apical progenitor (AP) maintenance. Together, our findings support a model in which lineage-specific / evolutionarily young transcription factor become integrated into a conserved developmental gene regulatory network, generating novel NPC dynamics during primate corticogenesis.

        Speaker: Cesar Mateo Bastidas Betancourt (German Primate Center)
    • 13:00 → 14:00
      Lunch 1h Restaurant

      Restaurant

      Harnack-Haus
    • 14:00 → 15:15
      Lecture: Randomised methods for handling large genomic datasets 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Tal Weizman Shapira

      Abstract
      In this lecture, I will talk about the basics of randomized methods for handling large genomic datasets. Namely, those needed for probabilistic counting and minimizer-based analysis. In the second part of the lecture, I will talk about an application, Needle, to quickly assess the expression of genes in thousands of files.

      Speaker: Knut Reinert (Freie Universität Berlin)
    • 15:15 → 15:45
      Coffee break & Transfer to FU 30m Lobby

      Lobby

      Harnack-Haus
    • 15:45 → 18:40
      Workshop Analysis of big data using Needle 2h 55m FU Berlin

      FU Berlin

      Reinert Lab

      Speakers: Jonas Schulte-Mattler (Freie Universität Berlin), Lennard Heimann (Freie Universität Berlin)
    • 18:40 → 19:00
      Transfer to Harnack Haus 20m
    • 19:00 → 20:00
      Dinner 1h Restaurant

      Restaurant

      Harnack-Haus
    • 20:00 → 21:00
      Networking with speakers 1h Lobby

      Lobby

      Harnack-Haus
  • Tuesday 8 September
    • 09:15 → 10:30
      Research Talk: Mobile DNA: from biology to technology 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Johnny Davila-Sandoval

      Abstract
      Transposable elements are major drivers of genome evolution and gene regulation, shaping genome architecture, promoting genetic diversity, and enabling horizontal gene transfer across all domains of life. Beyond their fundamental biological importance, these mobile genetic elements have inspired some of the most powerful technologies for genome engineering. Among them, CRISPR-associated transposons (CASTs) represent a recently discovered class of bacterial mobile genetic elements that combine the programmable DNA-targeting capability of CRISPR-Cas systems with the ability of transposases to insert DNA into the genome. Unlike conventional CRISPR-associated nucleases, CASTs enable RNA-guided DNA integration without requiring double-strand DNA breaks at the target site, making them particularly attractive as next-generation genome editing tools.

      In this talk, I will introduce the biology of transposable elements, highlighting their roles in genome plasticity, evolution, and horizontal gene transfer, and discuss how mobile DNA has become a valuable source of biotechnological innovation. I will then present our research on the structural basis of RNA-guided transposition by CRISPR-associated transposons and show how cryo-electron microscopy, combined with biochemical analyses, has uncovered the molecular mechanisms governing their function. I will discuss how these structural insights advance our understanding of mobile DNA while providing a foundation for the rational development of programmable genome engineering technologies.

      Speaker: Dr Irma Querques (Max Perutz Labs Vienna)
    • 10:30 → 11:00
      Coffee break 30m Lobby

      Lobby

      Harnack-Haus
    • 11:00 → 11:40
      Selected student talk 4 40m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Johnny Davila-Sandoval

      Speaker: Aruna Arumugam (Freie Universität Berlin)
      • Stimulus-Responsive Protein and RNA Networks of the ASCC Complex 40m

        Abstract:
        The human activating signal co-integrator complex (ASCC) consists of the core subunits ASCC1, ASCC2, ASCC3, and TRIP4 and associates with additional proteins to regulate genome maintenance and gene expression processes. However, it remains unclear how the core ASCC interacts with auxiliary factors, how these interactions relate to nucleic acid binding, and how the complex responds to cellular stimuli. We therefore mapped TRIP4-associated proteins and RNAs using proximity labeling, cross-linking and analysis of cDNA (CRAC) analysis. We generated CRISPR/Cas9-engineered HeLa cells producing TurboID-tagged TRIP4 and used established Flp-In™ T-REx™ 293 cells producing Flag-tagged TRIP4 for interactome analyses by proximity labeling and co-immunoprecipitation coupled to mass spectrometry, in the presence or absence of 9-cis retinoic acid (RA). These analyses revealed that TRIP4 participates in interconnected networks involved in transcription, translation, RNA metabolism, and cytoskeletal organization, and that these interactions are remodeled following RA treatment. Integration of TRIP4 CRAC data with the protein interactome identified extensive overlap at RNA regulatory hubs, particularly within the eukaryotic translation initiation factor 4F complex. Collectively, these findings provide a framework for understanding how ASCC and its associated factors coordinate multiple cellular processes.

        Speaker: Aruna Arumugam (Freie Universität Berlin)
    • 11:45 → 13:00
      Research Talk: Decoding the molecular mechanisms of nuclear organization via multiplexed nanoscopy 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair:Johnny Davila-Sandoval

      Abstract:
      Super-resolution microscopy has transformed our ability to probe cellular organization at the molecular scale. Among these approaches, DNA-PAINT (Points Accumulation for Imaging in Nanoscale Topography) provides spatial resolution down to the single-protein level, supports straightforward multiplexing, and enables molecular counting. DNA-PAINT relies on fluorescently labeled oligonucleotides that diffuse and stochastically bind to their complementary “docking strands” attached to target molecules. This equilibrium of binding and unbinding is both well characterized and highly programmable in sequence space, making DNA-PAINT a powerful tool for advanced single-molecule microscopy applications.

      In the first part of my talk, I will introduce the core principles of DNA-PAINT and highlight recent developments in the field, including our own contributions. These include the design of a tailored microscope, ultraprecise kinetic measurements that enable robust multiplexing and counting, and new single-particle-tracking implementations.

      In the second part of my talk, I will outline my current biological focus: refining and applying the quantitative DNA-PAINT toolbox to uncover how nuclear organization shapes genome function. By visualizing protein–RNA–DNA interactions with nanometer precision, we aim to understand the molecular principles that govern nuclear compartmentalization, many of which are linked to liquid-liquid phase separation. This pursuit has led us to the interface with electron microscopy, where integrating molecular mapping with ultrastructural context opens new opportunities for in situ structural biology.

      Speaker: Dr Johannes Stein (MPIMG)
    • 13:00 → 14:00
      Lunch 1h Restaurant

      Restaurant

      Harnack-Haus
    • 14:00 → 14:15
      Transfer to MPIMG 15m
    • 14:15 → 16:15
      Workshop From Raw Data to 3D Structure: A Practical Introduction to Cryo-EM – Part 1 2h SR1 (MPIMG)

      SR1

      MPIMG

      Abstract:
      Cryo-electron microscopy enables the structural analysis of biological macromolecules at high resolution, but successful experiments require an understanding of both experimental design and data processing.

      This workshop provides an accessible introduction to single-particle cryo-EM as well as cryo-electron tomography.
      The first part covers essential concepts for planning and evaluating cryo-EM experiments, including sample and grid requirements, vitrification, image formation, electron dose, contrast transfer function and data quality.

      In the practical part, we will work together to process a prepared mock dataset designed to complete the major processing steps within the workshop.

      By following the complete workflow from raw data to three-dimensional reconstruction, participants will gain a practical overview of cryo-EM data processing and learn how experimental choices and data quality affect the final structure.

      Speaker: Dr Thiemo Sprink (Max-Delbrück-Centrum für Molekulare Medizin)
    • 14:15 → 16:15
      Workshop Imaging Genetics - Light Microscopy – Part 1 2h SR2 (MPIMG)

      SR2

      MPIMG

      Abstract
      The workshop will introduce accessible assays that, through creativity and careful experimental design rather than sophisticated instrumentation, provide valuable insights into genetics. The goal is to explain, inspire and equip participants with light-microscopy-based approaches that can be readily implemented in their own labs.

      At the MPIMG we understand every microscopy application as a complete workflow. Therefore, the workshop will cover the entire process from experimental design, sample preparation, and data acquisition but will focus on image analysis and simple statistics.

      The planned topics will handle chromatin organization, mitotic indexing, spatial localization of genes, single RNA molecules, proteins, and cellular compartments (like organelles, condensates, or nucleoli). We will compare tracking and tracing, and discuss whether epigenetic markers together in imaging approaches are sufficient to define any cellular states.

      Speaker: Dr René Buschow (MPIMG)
    • 16:15 → 16:45
      Coffee break 30m Lobby (MPIMG)

      Lobby

      MPIMG
    • 16:45 → 18:45
      Workshop From Raw Data to 3D Structure: A Practical Introduction to Cryo-EM – Part 2 2h SR1 (MPIMG)

      SR1

      MPIMG

      Abstract:
      Cryo-electron microscopy enables the structural analysis of biological macromolecules at high resolution, but successful experiments require an understanding of both experimental design and data processing.

      This workshop provides an accessible introduction to single-particle cryo-EM as well as cryo-electron tomography.
      The first part covers essential concepts for planning and evaluating cryo-EM experiments, including sample and grid requirements, vitrification, image formation, electron dose, contrast transfer function and data quality.

      In the practical part, we will work together to process a prepared mock dataset designed to complete the major processing steps within the workshop.

      By following the complete workflow from raw data to three-dimensional reconstruction, participants will gain a practical overview of cryo-EM data processing and learn how experimental choices and data quality affect the final structure.

      Speaker: Dr Thiemo Sprink (Max-Delbrück-Centrum für Molekulare Medizin)
    • 16:45 → 18:45
      Workshop Imaging Genetics - Light Microscopy – Part 2 2h SR2 (MPIMG)

      SR2

      MPIMG

      Abstract
      The workshop will introduce accessible assays that, through creativity and careful experimental design rather than sophisticated instrumentation, provide valuable insights into genetics. The goal is to explain, inspire and equip participants with light-microscopy-based approaches that can be readily implemented in their own labs.

      At the MPIMG we understand every microscopy application as a complete workflow. Therefore, the workshop will cover the entire process from experimental design, sample preparation, and data acquisition but will focus on image analysis and simple statistics.

      The planned topics will handle chromatin organization, mitotic indexing, spatial localization of genes, single RNA molecules, proteins, and cellular compartments (like organelles, condensates, or nucleoli). We will compare tracking and tracing, and discuss whether epigenetic markers together in imaging approaches are sufficient to define any cellular states.

      Speaker: Dr René Buschow (MPIMG)
    • 18:45 → 19:00
      Transfer to Harnack-Haus 15m
    • 19:00 → 20:00
      Dinner 1h Restaurant

      Restaurant

      Harnack-Haus
    • 20:00 → 21:00
      Networking with speakers 1h Lobby

      Lobby

      Harnack-Haus
  • Wednesday 9 September
    • 09:15 → 10:30
      Research Talk: Decoding the Architecture(s) of Core Promoters 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Zhihao Shao

      Abstract:
      Most genome annotations mark a transcription start site with a single arrow at a single genomic position. Measured directly, most core promoters do not initiate transcription from a single point: initiation positions form a distribution, and the shape of this distribution carries the signature of promoter architecture and functional specialisation. Cap Analysis of Gene Expression, and its newest variants, remain the highest-resolution, least biased and most quantitative way to measure that distribution at RNA polymerase II promoters.

      We shall start with an overview of what CAGE measures and what it does not: cap trapping, counts at individual start sites, artefacts that survive into published tracks, normalisation across libraries of unequal depth, and the clustering decisions that quietly determine every result built on top of them.

      Three bodies of work will be given as worked examples. In zebrafish development, base-resolution mapping uncovered two independent initiation codes overlapping on the same core promoters, and a wholesale change of initiation grammar at the maternal-to-zygotic transition that no lower-resolution assay could have detected. Promoter architecture itself turns out to be a functional classification: interquantile width, CpG content and dual initiation separate classes of promoter that behave differently across developmental time and vertebrate evolution. The analysis of housekeeping promoters bound downstream of the start site by the transcription factor YY1 shows that YY1 constrains promoter width to protect the reading frame and its ribosome-docking motif, linking transcriptional precision to translational fidelity.

      Deep learning models trained on human sequence now claim to predict where transcription starts directly. We will take a critical look at some of those claims: what these models get right, where they fail, and what should count as evidence when agreement with someone else's data is the only test available.

      We close with some pointers on using the CAGEr and related packages and the CAGEflow pipeline for promoter analysis and data integration, and how to combine them with agentic coding to preempt common pitfalls.

      Speaker: Prof. Boris Lenhard (Imperial College London)
    • 10:30 → 11:00
      Coffee break 30m Lobby

      Lobby

      Harnack-Haus
    • 11:00 → 11:40
      Selected student talk 5 40m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Zhihao Shao

      Speaker: Marcel Wittmund (Max Planck Institute for Molecular Genetics)
      • Sequence, structure, and evolution of nucleolar targeting in ribosomal proteins 40m

        Abstract:
        In eukaryotes, 80 ribosomal proteins must be efficiently targeted to the nucleolus, where they assemble with four rRNAs to form functional ribosomes. The nucleolus is one of the most extensively studied biomolecular condensates, characterized by a multilayered, liquid-like architecture. Yet, the sequence features that direct ribosomal proteins to the nucleolus, and how these features have evolved, remain incompletely understood.

        To address this, we developed a high-throughput peptide-based screening strategy to systematically quantify the subcellular localization of more than 1,000 peptides tiling all human and E. coli ribosomal proteins. Using pooled oligonucleotide synthesis, lentiviral delivery, and stable cell line generation, we expressed peptide libraries fused to a fluorescent reporter and quantified their localization by high-throughput imaging and automated image analysis.

        This approach recapitulates known nucleolar localization signals, agrees with computational predictions, and identifies previously uncharacterized nucleolar-targeting regions within ribosomal proteins. To define the sequence logic underlying these localization patterns, we combined statistical sequence-feature analysis with machine-learning approaches, including convolutional neural networks for motif and pattern discovery. These models reveal peptide features predictive of nucleolar enrichment and enable the identification of sequence patterns that are not captured by simple composition-based descriptors alone.

        We then mapped experimentally identified targeting regions across approximately 1.5 million ribosomal proteins from nearly 50,000 species to investigate the evolutionary history of nucleolar-targeting features. By integrating primary sequence analysis, CNN-derived patterns, conservation profiling, and structural mapping within the ribosome, we assess how nucleolar localization signals are conserved, expanded, or remodeled across the tree of life.

        Together, this work provides a systematic experimental and computational framework for decoding the molecular grammar of nucleolar targeting. Our results reveal how short peptide-encoded features contribute to ribosomal protein localization and suggest how these features have evolved in relation to ribosome architecture and nucleolar organization.

        Speaker: Marcel Wittmund (Max Planck Institute for Molecular Genetics)
    • 11:45 → 13:00
      Research Talk: Building the Engines of Expression: From Ribosome Production to Translational Regulation 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Zhihao Shao

      Abstract:
      Human cells employ a variety of mechanisms to regulate growth and proliferation. Crucial amongst these is the continuous and accurate production of ribosomes. In fact, the steady production of functional ribosomes is closely linked to sustaining life. This process, known as ribosome biogenesis, is a highly complex biochemical pathway and one of the most energy-intensive processes in the cell, requiring the concerted action of hundreds of assembly factors and ribosomal proteins. Defects in ribosome biogenesis can lead to ribosome hypo-proliferation and heterogeneity in ribosome production, adversely affecting cellular homeostasis and leading to pathogenesis. Such defects are known to give rise to the heterogeneous group of rare diseases known as ribosomopathies. In my talk, I will detail how human cells produce ribosomes, and how defects in their production lead to disease. Additionally, I will discuss other aspects of translational regulation, as well as emerging technologies in this field.

      Speaker: Dr Sameer Singh (Charité – Universitätsmedizin Berlin)
    • 13:00 → 14:00
      Lunch 1h Restaurant

      Restaurant

      Harnack-Haus
    • 14:00 → 14:15
      Transfer to MPIMG 15m
    • 14:15 → 16:15
      Poster session (odd numbers) 2h Lobby (MPIMG)

      Lobby

      MPIMG
      Poster Abstract Session 1
    • 16:15 → 16:45
      Coffee break 30m Foyer (MPIMG)

      Foyer

      MPIMG
    • 16:45 → 18:45
      Poster session (even numbers) 2h Lobby (MPIMG)

      Lobby

      MPIMG
      Poster Abstract Session 2
    • 18:45 → 19:00
      Transfer to Harnack-Haus 15m
    • 19:00 → 20:00
      Dinner 1h Restaurant

      Restaurant

      Harnack-Haus
    • 20:00 → 21:00
      Networking with speakers 1h Lobby

      Lobby

      Harnack-Haus
  • Thursday 10 September
    • 09:15 → 10:30
      Research Talk: Decoding the Spatial Layer of Cellular Regulation with a DNA Toolbox 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Celina Adalem

      Abstract:
      Cellular function is shaped not only by molecular composition but also by the spatial organization of molecules within cells. Our research seeks to make this spatial layer of biology directly measurable by integrating chemistry, advanced microscopy, omics-scale molecular profiling, and computational analysis. Central to this effort is the use of DNA as a programmable engineering material for molecular tagging, barcoding, amplification, and in situ or ex situ readouts.

      We have developed several technologies for spatial molecular analysis, including SABER for highly multiplexed and sensitive imaging of RNA and proteins, Light-Seq for linking optical selection to transcriptome-wide sequencing of defined cells and subcellular regions, and ProPER, a new proximity detection approach for visualizing molecular interactions and reporting regulatory states in situ.

      Using this toolkit, we investigate how the spatial organization of RNAs and proteins regulates cellular function in health and disease, with a particular focus on neurons, where extreme compartmentalization imposes unique constraints on gene regulation. By directly measuring molecular localization, interactions, and regulatory states in cells and tissues, we aim to uncover organizational principles that shape physiology and contribute to disorders such as neurodegeneration and cancer.

      Speaker: Dr Sinem Saka (EMBL Heidelberg)
    • 10:30 → 11:00
      Coffee break 30m Lobby

      Lobby

      Harnack-Haus
    • 11:00 → 11:40
      Selected student talk 6 40m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Celina Adalem

      Speaker: Emilia Rezzolla (Max Planck Institute for Molecular Genetics)
      • Analyzing the response of gene regulatory DNA elements to the transcription factor Oct4 using the tunable CRISPRi system CasTuner 40m

        Abstract:
        Since the discovery of the bacterial antiviral defense system CRISPR-Cas9 and its revolutionary impact on genetic engineering, significant efforts have been made to develop precise and efficient tools capable of targeting specific DNA sequences in living cells. Various CRISPR-Cas9 modifications have been engineered to enable targeted double-strand DNA breaks, base editing, gene knockdown (KD), transcriptional regulation, and epigenome modification.
        Despite these advances, a common limitation of existing CRISPR-based tools is their binary activity states, active or inactive, without accounting for the dose-dependent nature of many cellular processes, such as sex-chromosome dosage compensation and haploinsufficiency, where the abundance of specific factors critically influences downstream activities.
        To address this limitation, the CRISPR-based tool CasTuner was developed. This system allows in vivo modulation of protein abundance through degron-based regulation, enabling precise control over target protein knockdown, including transcription factors (TFs). By modulating TF abundance, CasTuner can alter downstream gene expression patterns, offering a novel approach to studying transcriptional regulation and the role of TFs in their native cellular contexts.
        One exemplary TF of particular interest is Oct4, known to drive differential gene expression depending on its cellular abundance. This thesis investigates the functionality of CasTuner by analyzing the transcription of reporter genes driven by computationally predicted cis-regulatory elements (CREs), which are targeted by Oct4 and sensitive to Oct4 dosage perturbations. The findings aim to demonstrate the innovative utility of CasTuner and emphasize the critical importance of considering how TF dosage shapes gene expression outcomes.

        Speaker: Emilia Rezzolla (Max Planck Institute for Molecular Genetics)
    • 11:45 → 13:00
      Research Talk: Combining experiments and computation to understand the epigenetic regulation in single cells 1h 15m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Chair: Celina Adalem

      Abstract:
      Cellular gene expression programs are driven by multiple layers of epigenetic regulation and mapping them could help us better understand how to manipulate cell behaviour. Recently, experimental assays have evolved in sensitivity. This now enables us to map the events such as DNA accessibility, methylation and histone modifications in individual cells, resolving tissue heterogeneity. However, single-cell genomics also brings new challenges, emerging from noisy measurements, high dimensionality, and assay-specific biases. I will present the computational tools we have developed to tackle these challenges and discuss how we are leveraging the signal from these emerging assays to understand the epigenetic regulation of embryonic development.

      Speaker: Dr Vivek Bhardwaj (Utrecht University)
    • 13:00 → 14:00
      Lunch 1h Restaurant

      Restaurant

      Harnack-Haus
    • 14:00 → 15:30
      Workshop Building Your Career Strategy for Academia & Beyond 1h 30m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      A career strategy is not a fixed plan. It starts with understanding what drives you, what you are good at, and what you need to do your best work. In this interactive workshop, participants will reflect on their personal definition of success, explore how they want to contribute through their work, and learn how to communicate their skills and strengths to different audiences. We will also examine how purposeful networking and visibility can help create opportunities in academia and beyond. The session combines individual reflection, group work, practical input, and Q&A.

      Speakers: Dr Birte Seffert (GSO – Guidance, Skills & Opportunities for Researchers e. V.), Tobias Renner (GSO – Guidance, Skills & Opportunities for Researchers e. V.))
    • 15:30 → 16:00
      Coffee break 30m Lobby

      Lobby

      Harnack-Haus
    • 16:00 → 17:30
      Workshop Building Your Career Strategy for Academia & Beyond 1h 30m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      A career strategy is not a fixed plan. It starts with understanding what drives you, what you are good at, and what you need to do your best work. In this interactive workshop, participants will reflect on their personal definition of success, explore how they want to contribute through their work, and learn how to communicate their skills and strengths to different audiences. We will also examine how purposeful networking and visibility can help create opportunities in academia and beyond. The session combines individual reflection, group work, practical input, and Q&A.

      Speakers: Dr Birte Seffert (GSO – Guidance, Skills & Opportunities for Researchers e. V.), Tobias Renner (GSO – Guidance, Skills & Opportunities for Researchers e. V.))
    • 17:35 → 19:00
      Team building 1h 25m Hahn-Hörsaal

      Hahn-Hörsaal

      Harnack-Haus

      Ihnestrasse 16-20 - 14195 Berlin, Germany

      Moderation: Tommaso Stentella

    • 19:00 → 20:00
      Dinner 1h Restaurant

      Restaurant

      Harnack-Haus
    • 20:00 → 21:00
      Networking with speakers 1h Lobby

      Lobby

      Harnack-Haus
  • Friday 11 September
    • 09:00 → 10:20
      Selected student talks 7 & 8 1h 20m SR1 (MPIMG)

      SR1

      MPIMG

      Chair: Sara Lopez Ruiz de Vargas

      Speakers: Beate Bergman (University Medical Center Göttingen), Ekin Deniz Aksu (Vingron Lab (MPIMG))
      • How hungry proteins battle mutational meltdowns: ULK1-dependent mitophagy and somatic purifying selection of the mitochondrial genome 40m

        Abstract:
        Mitochondrial DNA (mtDNA) is present in multiple copies per cell and is vulnerable to mutation accumulation due to relaxed segregation, limited recombination, and continuous replication. Nevertheless, observed deleterious mtDNA mutations are often maintained below pathogenic levels, suggesting active purifying selection mechanisms. While our group has previously established mitophagy as a selection mechanism in the germline, its role in somatic tissues remains unclear. To address this, we use two complementary mouse models: the mt-tRNA-Ala C5024T mouse, carrying a single pathogenic mtDNA mutation at high heteroplasmy, and the PolG mutator mouse, which accumulates many random low-level mtDNA mutations due to impaired polymerase proofreading. These models were combined with knockout of ULK1, a key autophagy initiation factor. Aged mice showed tissue-specific phenotypes, particularly altered spleen and heart weights. However, RT-qPCR analysis of mitochondrial stress and inflammatory markers in C5024T tissues did not indicate a strong canonical mitochondrial stress response.
        In line with our hypothesis, loss of ULK1 increased C5024T heteroplasmy across somatic tissues, supporting a role for ULK1-dependent mitophagy in limiting expansion of a pathogenic mtDNA mutation. Surprisingly, the opposite was observed in PolG mutator mice, where ULK1 knockout reduced mtDNA variant burden. We propose that in a background with many low-frequency mutations, increased activation of mitophagy may preferentially remove functional mtDNA molecules, reducing the absolute number of functional mitochondria below a physiological threshold. Proteomics of mutator spleens pointed toward compensatory LC3-associated mitophagy and altered erythropoiesis, prompting flow cytometric analysis. Mutator ULK1 knockout mice showed mitochondrial retention in circulating erythrocytes and increased erythrocyte abundance in spleen, suggesting that aggravated splenomegaly is partly caused by a mitochondrial anemia-like phenotype.
        Together, these data suggest that ULK1-dependent mitophagy influences somatic mtDNA selection, but its outcome depends on the mutational landscape: protective against high-heteroplasmy pathogenic mutations, yet disadventageous in complex low-level mutator backgrounds.

        Speaker: Beate Bergman (University Medical Center Göttingen)
      • Corgi: context-aware sequence-to-function model of human gene regulation 40m

        Abstract:
        Sequence-to-function models have been very successful in predicting gene expression, chromatin accessibility, and epigenetic marks from DNA sequences alone. However, current models have a fundamental limitation: they cannot extrapolate beyond the cell types and conditions included in their training dataset. Here, we introduce Corgi, a new context-aware sequence-to-function model that overcomes this limitation by integrating DNA sequence and trans-regulator expression to predict chromatin accessibility, histone modifications and gene expression coverage, even in held-out cell types. Trained on a diverse set of bulk and single cell human datasets, Corgi achieves state-of-the-art performance in joint cross-sequence and cross-cell type epigenetic track prediction. Additionally, we present an advanced model version, Corgi+, which is state-of-the-art in imputation of epigenetic tracks using only RNA-seq data. We further show that Corgi learns key cell type-specific trans-regulators in a zero shot manner, and it can predict genomic variant effects in held-out cell types.

        Original publication: https://www.nature.com/articles/s41467-026-75527-2

        Speaker: Ekin Deniz Aksu (Vingron Lab (MPIMG))
    • 10:20 → 10:45
      Coffee break 25m Lobby (MPIMG)

      Lobby

      MPIMG
    • 10:45 → 12:00
      Research Talk: From multi-omics to mechanisms of chromatin regulation 1h 15m SR1 (MPIMG)

      SR1

      MPIMG

      Chair: Sara Lopez Ruiz de Vargas

      Abstract:
      Gene expression is regulated through a hierarchy of interconnected processes spanning multiple biological scales, from signaling to transcription factor binding, to enhancer activity, gene regulatory networks and ultimately stable cell identities. In this lecture, I will discuss how we leverage chromatin accessibility in combination with transcriptomics and proteomics as a versatile framework for dissecting regulatory mechanisms across these scales. Using examples from neuronal differentiation and human stem cell models of neurodevelopmental disorders, I will illustrate how multi-omic integration enables inference of transcription factor activity, identification of dynamic epigenetic memory during cell fate transitions, and reconstruction of enhancer-mediated gene regulatory networks that reveal mechanisms not apparent from differential gene expression alone.

      Speaker: Dr Daria Bunina (Max-Delbrück-Centrum für Molekulare Medizin)
    • 12:00 → 13:00
      Lunch 1h Canteen (MPIMG)

      Canteen

      MPIMG
    • 13:00 → 14:00
      Feedback & Closing 1h SR1 (MPIMG)

      SR1

      MPIMG