Detailed Scientific Research & Implementation Strategies
Marina Elez is a pioneer in developing single-cell approaches to observe replication errors in real-time, particularly in E. coli. Her group utilizes varied microscopy techniques coupled with microfluidics to track the fate of specific replication errors. By fluorescently labeling components of the mismatch repair (MMR) system, she has been able to visualize the dynamics of error detection and repair. This is significant because it moves beyond population-level sequencing data to show exactly when and where errors arise and whether they are repaired or become fixed mutations.
Her research has revealed that there is a "temporal constraint" on repair; often, mutations arise not because errors aren't detected, but because they aren't repaired in time before the next replication cycle. This single-cell view also uncovered heterogeneity in repair capacity among genetically identical cells.
Implementing this knowledge in your project could involve adopting similar time-lapse microscopy approaches to track DNA damage markers (like gamma-H2AX or specific repair protein foci) in real-time. If you work with yeast, adapting her microfluidic + fluorescent reporter strategies could allow you to distinguish between successful repair events and "failed" repair that leads to mutations or cell death, providing a more granular readout than simple viability assays.
Workshop: Single-cell level analysis
This practical course bridges her lecture's theory with reality. You will learn the computational and analytical workflows her lab uses to track replication errors in real-time. Expect to analyze time-lapse microscopy datasets generated from microfluidic chips, learning how to segment cells and track fluorescent MMR foci over time—essential skills if you want to execute the implementation strategy above!
Raphaël Mourad specializes in applying machine learning and deep learning to genomics, with a specific focus on predicting DNA Double-Strand Breaks (DSBs). His work, such as the widely cited research on predicting DSBs using epigenome marks, demonstrates that the location of breaks is not random but strongly correlated with chromatin context. He uses advanced computational models (including CNNs and DNABERT) to integrate varying data types—chromatin accessibility, transcription activity, and 3D genome data—to predict where the genome is most fragile.
His research emphasizes that "context is key." DSB formation isn't just about sequence motifs but about the functional state of the chromatin. He has developed tools to classify and predict these events with high accuracy by feeding large-scale epigenomic datasets into predictive models.
For your project, his methods offer a roadmap for integrating your own omics data. If you have ChIP-seq and DSB maps (like BLISS or ChIP of damage markers), you could apply similar ML frameworks to predict likely breakage sites under your specific stress conditions. Learning his workflows for data preprocessing and model training would enable you to build predictive models for your yeast or human cell systems.
Workshop: Machine learning applications
This session is a hands-on extension of his lecture. You will dive into building or utilizing Machine Learning models (such as CNNs) to predict DSB formation from epigenomic data. This is where you actually learn the code and frameworks required to process 3D genomics data and train predictive networks.
Dana Branzei represents a highly specialized expertise in structural verification of replication intermediates. Her lab combines genetics with physical handling of DNA molecules using 2D agarose gel electrophoresis and Transmission Electron Microscopy (TEM). While sequencing gives "lists" of reads, Branzei's work provides the "shape" of the DNA. She visualizes complex structures like reversed forks (chicken-foot structures), hemicatenanes, and recombination intermediates.
Her research focuses on the "DNA Damage Tolerance" (DDT) pathways. By using TEM, she can physically see single-stranded gaps and the topology of sister chromatid junctions, proving mechanistically how a cell handles a stalled fork.
Implementation in your project would be transformative for validating mechanistic hypotheses. If you propose that a certain mutant accumulates reversed forks, learning to interpret 2D gels or collaborating on EM samples would provide the "smoking gun" evidence. Even understanding the logic of her analysis will help you better interpret your own replication profiling data.
Judith Mine-Hattab investigates the biophysics of nuclear organization, specifically how liquid-liquid phase separation (LLPS) drives the early DNA damage response. Her work on FUS shows that this protein undergoes phase separation to rapidly compartmentalize DNA damage sites. This allows for the rapid concentration of repair factors while dynamically regulating their access.
She uses advanced single-molecule tracking (SMT) to measure the diffusion of molecules inside vs. outside these foci, revealing that inside these droplets, molecules are not just trapped; they exhibit distinct diffusion properties that facilitate efficient biochemistry.
For your project, her techniques offer a way to measure the quality of your repair foci, not just their number. By implementing single-particle tracking analysis, you could determine if your protein of interest is dynamically exchanging at damage sites or if it forms rigid aggregates.
Workshop: FreeTrace and BI-ADD, SPT software
Her lecture explains why measuring FUS phase separation matters; this workshop shows you how. You will learn to use `FreeTrace` and `BI-ADD` to extract critical biophysical parameters (diffusion constants, confinement radii) from Single-Molecule Tracking (SPT) data. If you ever plan to do live-cell imaging of your own repair proteins, these software tools are how you prove they are biologically active.
Nicola Crosetto is a co-developer of the BLISS (Breaks Labeling In Situ and Sequencing) method, a high-sensitivity technique for mapping DSBs genome-wide. Unlike ChIP-seq for gamma-H2AX which covers broad domains, BLISS captures the exact nucleotide resolution of the break ends.
His recent work has expanded to sBLISS (in suspension), making the protocol scalable. He generates massive datasets connecting spontaneous DNA damage to structural variants (translocations) in cancer, correlating the frequency of breakage at a locus with its propensity to translocate.
Learning his quantitative workflows allows you to move beyond "more" or "less" damage to "where" and "why." If you can implement BLISS in your lab, you could map the exact sites of instability induced by your mutant of interest. His analysis pipelines for filtering "biological noise" are essential for anyone working with DSB-seq data.
Workshop: BLISS data processing & analysis
Since Nicola developed the BLISS method, his lecture proves its immense power. His practical course is the bioinformatics tutorial you need to actually use it. You will take raw BLISS sequencing reads, filter out biological noise, and map exact breaking hotspots onto the genome computationally. This translates his high-level genomic fragility concepts into code you can run.
Chunlong Chen focuses on the "replication program"—the strict timing and spatial organization of origin firing. Using single-cell copy number variations (scCNV) to infer replication states, he builds "atlases" of replication timing that reveal cell-to-cell heterogeneity.
His team develops computational tools to disentangle replication signals from somatic copy number alterations in cancer, linking this timing data with gene expression and chromatin structure.
If you have single-cell sequencing data, his computational methods (e.g., using scCNV to infer S-phase progression) would be invaluable. Instead of sorting cells by FACS, you could computationally sort them by replication progress, providing higher resolution phenotypic analysis of your replication stress models.
Workshop: Replication dynamics workshop
While his lecture focuses on the 'replication program' and cell-to-cell heterogeneity, this practical course will teach you the specific computational tools needed to map origin firing and replication progression. You will learn how to disentangle replication signals from DNA sequencing data, turning raw reads into the beautiful atlases of replication timing he presents.
This lecture combines experimental yeast genetics (Teixeira) with applied mathematics (Doumic). Their research addresses the "counting mechanism" of senescence, showing that the shortest telomere triggers arrest, not the average. Doumic develops mathematical models that predict the distribution of telomere lengths over generations.
Teixeira's experimental work in yeast uses single-cell lineages and microfluidics to track the onset of senescence. By combining this with mathematical models, they can distinguish between gradual shortening and sudden, stochastic loss events.
This is a prime example of "Quantitative Biology." For your project, their work teaches you how to model population dynamics. If you observe a phenotype that appears "stochastic" or variable, their modeling approach helps you test if this variability is an intrinsic property of the mechanism or due to external noise.
Workshop: Telomere analysis workshop
This workshop perfectly embodies the Qlife Winter School's goal: bridging biology and math. Teresa provides the experimental yeast senescence data, and Marie provides the mathematical framework. In the practical, you will learn how to apply these mathematical models to the biological data, calculating the stochastic loss events and predicting population distributions over generations.
Angela Taddei studies the search for homology, the "needle in a haystack" problem of DNA repair. Her group uses live-cell imaging to track the movement of damaged DNA (marked by Rad52 or Rad51) as it scans the nucleus for a template. Her key finding is that upon damage, chromatin becomes more mobile, increasing the volume it explores.
Recent work involves tracking Rad51 filaments in vivo, describing how they extend and contract via "intersegmental transfer", quantifying parameters like "radius of confinement" and MSD.
From her lecture, you can extract specific image analysis protocols for tracking loci. Calculating MSD and confinement radii is the standard for proving "chromatin mobility." Implementing this allows you to quantitatively claim that a repair protein "mobilizes" chromatin, bridging microscopy to polymer physics models.
Corinne Grey focuses on meiotic recombination, specifically the structural organization of the chromosome axis. Her research highlights that recombination happens within loop-axis structures, where "hotspots" are tethered to the chromosome axis by proteins like IHO1. This ensures that breaks formed in the loop are effectively processed by the repair machinery located on the axis.
She maps these interactions using high-resolution genomics and investigates the "feedback loops" where crossover formation inhibits nearby breaks (interference).
Even if you study mitotic cells, the concept of "loop tethering" is relevant (e.g., cohesion-mediated loops). Her methods for mapping protein-DNA topology (ChIP-seq relative to Hi-C loops) could help you map your repair proteins relative to architectural markers like Cohesin or CTCF.
Martin Taylor investigates the "dark side" of DNA repair: mutagenesis. His work focuses on DNA Damage Tolerance (DDT) and translesion synthesis (TLS). He uses "mutational signatures" to identify which polymerase or pathway was active, revealing how replication and transcription coupled repair shape the mutational landscape.
He studies "lesion segregation," where a DNA adduct isn't repaired but passed to a daughter cell, inducing mutations in subsequent generations. His forensic reconstructions from whole-genome sequencing decode exactly what happened at the replication fork.
If you have sequenced genomes of mutant strains, you can use frameworks (R packages like MutationalPatterns) to decompose mutation lists into signatures. This tells you if a deletion causes a specific defect simply by reading the "fingerprint" of the mutations left behind.
Workshop: Mutation clusters and asymmetry
This selected practical course immediately empowers you to execute his lecture's concepts. You will use R/computational frameworks to trace mutation signatures and cluster patterns from whole-genome sequencing datasets, learning how to decipher the fingerprints of specific DNA damage mechanisms.
Aurèle Piazza bridges cohesin biology and DNA repair. His work demonstrated that cohesin actively constrains the homology search to the cis chromatid. Utilizing loop extrusion models, he showed that cohesin creates a "local" search space, promoting specific repair with the sister chromatid while suppressing "illegitimate" recombination.
His methods rely heavily on Hi-C combined with induced breaks. He quantifies how contact frequency changes upon damage induction, allowing him to propose mechanical models of DNA spatial reorganization during repair.
The Hi-C analysis workshop is his direct implementation. You will learn to process Hi-C matrices to identify loops and compartments. Implementing this means checking if the "search space" is altered in your mutants using Hi-C to check for loss of loop extrusion or compartment integrity.
Workshop: Analysis of Hi-C data
This hands-on workshop is how you actually measure the 'search space' he discusses in his lecture. You will learn to process raw Hi-C contact matrices computationally, identifying loops, domains, and compartments—an essential skill to link genome architecture to your own biological hypotheses.
Leonid Mirny is a theoretical physicist known for fundamentally shifting our understanding of the 3D genome. His lab developed the "loop extrusion" model of chromatin organization and studies how polymer physics dictate the search processes of repair proteins. In the context of "partner search," his simulations show how diffusion in a polymer melt is constrained by the chain.
His work models the search time—how long it takes for locus A to find locus B—predicting that "1D sliding" combined with "3D hopping" speeds this up.
The implementation here is Polymer Simulation. For your project, this is powerful for hypothesis testing. If you think two enhancers interact, you can model the polymer to see if that interaction is physically probable given loop constraints. It helps you critically quality-check Hi-C data against physical reality.
Workshop: Modeling search: polymer dynamics
Moving from high-level physics to actual commands, this practical will train you in computational polymer modeling. You will run physical simulations to model how chromatin folding constraints impact protein target search times, turning theoretical biophysics into reproducible code.
This duo represents the cutting edge of High-Content Screening and AI-driven image analysis. Hana Polasek-Sedlackova uses quantitative microscopy to image thousands of cells to understand MCM helicase loading. Paolo Fagherazzi specializes in the AI workflows. Their joint work focuses on removing human bias from image quantification.
They use tools that segment nuclei, detect foci, and measure signal intensity automatically across huge datasets, emphasizing "single-cell" statistics (plotting distributions rather than means) to reveal rare subpopulations.
Instead of manually counting foci in 50 cells, you would implement their Automated Image Analysis Pipelines (e.g., CellProfiler, Cellpose). This allows you to analyze 10,000 cells, giving you the statistical power to detect subtle phenotypes that manual quantification would miss.
Workshop: Microscopy image analysis
Your chosen session is purely hands-on AI bio-imaging! You will practice taking massive microscopy images and using automated pipelines (like Cellpose/AI variants) to segment nuclei and flawlessly quantify damage repair foci without human bias. This directly applies their high-content screening concepts to your physical benchwork data.
Terence Strick pioneers single-molecule biophysics, employing nanomanipulation techniques like magnetic tweezers to observe the real-time dynamics of DNA repair machineries. His laboratory constructs physical models of how massive multi-protein complexes assemble to execute transcription-coupled repair.
Patrick Charnay is a central figure in developmental genetics, integrating molecular biology and genetics to understand cellular differentiation. Their combined expertise underpins the structural organization of this quantitative biology retreat.