Upcoming Events
Theory Lunch with Suri Vaikuntanathan "Limits on the computational expressivity of non-equilibrium biophysical processes"
Limits on the computational expressivity of non-equilibrium biophysical processes
Suri Vaikuntanathan | Department of Chemistry at University of Chicago
Abstract
Many biological decision-making processes can be viewed as performing a classification task over a set of inputs, using various chemical and physical processes as "biological hardware". In this context, it is important to understand the inherent limitations on the computational expressivity of classification functions instantiated in biophysical media. Here, we model biochemical networks as Markov jump processes and train them to perform classification tasks, allowing us to investigate their computational expressivity. We reveal several unanticipated limitations on the input-output functions of these systems, which we further show can be lifted using biochemical mechanisms like promiscuous binding. We analyze the flexibility and sharpness of decision boundaries as well as the classification capacity of these networks. Additionally, we identify distinctive signatures of networks trained for classification, including the emergence of correlated subsets of spanning trees and a creased "energy landscape" with multiple basins. Our findings have implications for understanding and designing physical computing systems in both biological and synthetic chemical settings.
Theory Lunch with Zev Gartner "Order from disorder: self-organization in development and disease"
Order from disorder: self-organization in development and disease
Zev Gartner | School of Pharmacy at UCSF
Abstract
Tissues are wildly complex, with properties that emerge from the interactions of large numbers of cells comprising a dizzying number of heterogeneously expressed gene products. The tools of genomics and big data are increasingly viewed as the solution to understanding this complexity. While the utility of these approaches are undeniable, we are exploring a parallel approach. Using bioengineering tools and human mammary organoids as a model system, we provide evidence that the conceptual tools of equilibrium statistical mechanics can provide surprisingly accurate predictions of steady-state phenomena at the tissue scale from only three measurable parameters — an active surface energy, the magnitude of active mechanical fluctuations, and a configurational entropy associated with composing a tissue from different populations of cells. From these measurements, we predict the average structure of a tissue across a range of conditions as well as its microscopic variability. This conceptual formalism also provides insight into how changes to these parameters can drive corresponding changes in tissue structure, for example during development and breast cancer progression. I will discuss some assumptions and limitations of this approach, possible extensions to other systems, and the potential to understand other emergent properties of tissues such as cell plasticity and structural phase transitions.
Theory Lunch with Benjamin Allen "Is it time to stop fighting about cooperation?"
Is it time to stop fighting about cooperation?
Benjamin Allen | Emmanuel College, Boston
Abstract
The evolution of cooperation has been a major question in evolutionary theory since Darwin. It has also been the focus of intense disputes over the merits of different methodological approaches. I will review the four major theoretical approaches to the evolution of cooperation: multilevel selection, inclusive fitness, population genetics, and evolutionary game theory. Crucially, these approaches differ in the level at which the explanation applies, and in how they relate to the notion of "actor". I will examine the question of whether these are mathematically equivalent, and critically evaluate previous claims of such equivalence. I will then outline a new mathematical approach to the evolution of cooperation and other social behavior, which synthesizes and extends the four existing frameworks. In this new approach, the notion of "actor" is extended from individuals to collectives, and the outcome of evolution reflects an aggregation over individual and collective interests across scales.
Theory Lunch with Steven Orzack "Science needs philosophy, but what philosophy?”
Science needs philosophy, but what philosophy?
Abstract
Forthcoming
Theory Lunch with Leonor Saiz "The reasonable and unreasonable effectiveness of equilibrium statistical mechanics in gene regulation"
The reasonable and unreasonable effectiveness of equilibrium statistical mechanics in gene regulation
Leonor Saiz | Department of Biomedical Engineering at UC Davis
Abstract
Gene regulation exemplifies a duality wherein equilibrium statistical mechanics effectively describes quasi-equilibrium behaviors, yet systems maintain functionality under significant perturbations. Time scale separation facilitates quasi-equilibrium conditions, enabling predictive modeling of regulatory mechanisms. However, the persistence of regulatory function during substantial disturbances, such as DNA replication, challenges equilibrium assumptions. This robustness can emerge from compensatory mechanisms, such as transcription factor coordination and DNA modifications, that mitigate transient non-equilibrium effects. This talk will examine how these strategies complement quasi-equilibrium dynamics, revealing the strengths and limits of equilibrium models in explaining gene regulation resilience.
Theory Lunch with Andrea Liu "Complex biological function as an inverse input-output problem"
Complex biological function as an inverse input-output problem
Andrea Liu | Dept of Physics and Astronomy at Univ. of Pennsylvania School of Arts & Sciences
Abstract
In neural networks, parameters that govern interactions between nodes are tuned to obtain desired input-output relations. I argue that it is useful to think of systems with biological function as having effective interactions that are tuned to achieve the biological function. Such systems can be viewed as members of a large class of systems that I call "tunable matter." In some cases, biological evolution has tuned the interactions, but in other cases, systems tune their effective interactions by local rules in order to maintain function on time scales much shorter than evolutionary ones. I argue that tunable matter provides a unifying conceptual framework for understanding the emergence of collective function in a wide range of living systems.
Theory Lunch Lucy O'Brien "How flux through the cell lifecycle enables organ adaptation"
How flux through the cell lifecycle enables organ adaptation
Lucy O'Brien | Dept of Molecular and Cellular Physiology at Stanford University School of Medicine
Abstract
Forthcoming
Theory Lunch with Saad Bhamla "Movement sans ATP and learning sans neurons in curious ciliates"
Movement sans ATP and learning sans neurons in curious ciliates
Saad Bhamla | School of Chemical and Biomolecular Engineering
Abstract
Forthcoming
Pizza Talks | Max Brambach
On Zoom and in WAB 563
Pizza Talks: Max Brambach, Oyler-Yaniv Lab
Talk Title: Uncovering how immune-niche organization regulates T-cell expansion
Pizza Talks | Kotaro Fujimaki
On Zoom and in WAB 563
Pizza Talks: Kotaro Fujimaki, Lahav Lab
Talk Title: DNA damage checkpoints balance a tradeoff between diploid- and polyploid-derived arrest failures
Pizza Talks | Xiang Deng
On Zoom and in WAB 563
Pizza Talks: Xiang Deng, Mitchison Lab
Talk Title: Beyond Structure: The Regulatory Role of Tubulin
Theory Lunch with Sahin Naqvi "Tradeoffs between transcriptional robustness and dosage sensitivity in development"
Tradeoffs between transcriptional robustness and dosage sensitivity in development
Sahin Naqvi | Division of Gastroenterology at Boston Children's Hospital
Abstract:
Developmental systems must balance robustness to perturbations with sensitivity to a broad range of regulatory inputs, a tradeoff that underpins both stability and adaptability. Transcription factors (TFs) embody this tension; while development is broadly buffered against quantitative changes in TF dosage, human genetics has revealed exquisite sensitivity, as exemplified by haploinsufficiency and dosage-sensitive disease phenotypes. I will discuss how, by precisely quantifying the effects of craniofacial TF dosage on molecular and cellular phenotypes, we arrived at a model that reconciles dosage sensitivity with robustness. In this model, most SOX9-dependent regulatory elements and genes are robust to quantitative changes in TF dosage, while others act as sensitive effectors, amplifying small changes in dosage into phenotypic diversity or vulnerability. I will discuss the generalizability of this model and some evolutionary implications.
Department Seminar | Jennifer Smith
Director, ICCB-Longwood Screening Facility
Talk Title: “Resources beyond screening at ICCB-Longwood Screening Facility”
Food to be Provided
Theory Lunch w/ Richard Watson “Learning, memory and adaptation before evolution”
Richard Watson | Institute for Life Sciences/ Department of Computer Science (Agents, Interaction and Complexity group), University of Southampton
Title: Learning, memory and adaptation before evolution
Abstract: Forthcoming
Theory Lunch w/ Stephen Wolfram “Computational irreducibility and minimal models of biology"
Stephen Wolfram | Founder & CEO of Wolfram Research
Title: Computational irreducibility and minimal models of biology
Abstract: Forthcoming
Pranam Chatterjee, Duke University
Title: Programmable Protein Therapeutics via Generative Language Models
Abstract:
CRISPR has revolutionized biotechnology by enabling the simple design of guide RNAs to target and edit almost any DNA sequence. By developing new generative protein design algorithms, my hybrid lab focuses on extending this CRISPR-like programmability to proteins and other key molecules. In this talk, we will first delve into our algorithms that design binders to undruggable proteins, such as those driving pediatric cancers (alveolar rhabdomyosarcoma and Ewing’s sarcoma) and neurodegenerative diseases (Huntington’s and Alexander Disease). Our generative language models, including SaLT&PepPr, PepPrCLIP, and PepMLM, design short binding peptides from target sequence alone, with no dependence on stable 3D structures, and by fusing these "guide" peptides to E3 ubiquitin ligases, deubiquitinases, and other modifying enzymes, we have created a CRISPR-analogous system to edit these proteins. To be even more specific, we train isoform-specific targeting models such as PTM-Mamba for PTM-specific binding, FusOn-pLM for fusion oncoprotein-specific degradation, and moPPIt for motif-specific targeting of protein-protein interactions. Inspired by the power of language models, we further show how we can extend this programmability to DNA with our PAM-free CRISPR enzymes and our recent DPAC model, as well as heavy metals through our MetaLATTE algorithm and chemical pollutants, such as PFAS. Finally, we will explore our long-term goal of generating new cell states with model-designed proteins, highlighting our recent work on transcription factor-directed stem cell differentiation to ovarian cell types, such as granulosa cells and oogonia. By combining generative design with experimental engineering, our hybrid lab aims to translate these advances into practical applications for treating intractable diseases and addressing environmental challenges.
Theory Lunch w/Stacey Finley “Studying the tumor microenvironment as an ecosystem"
Stacey Finley | University of Southern California Viterbi School of Engineering
Title: Studying the tumor microenvironment as an ecosystem
Abstract: Forthcoming
Theory Lunch w/ Adam Palmer How combination therapy cures some cancers"
Adam Palmer | University of North Carolina
Title: How combination therapy cures some cancers
Abstract: Forthcoming
Theory Lunch w/Mohit Kumar Jolly "Low dimensionality of phenotypic space as an emergent property of biological regulatory networks"
Mohit Kumar Jolly | Biological Sciences Building Indian Institute of Science, Bengaluru, India
Title: Low dimensionality of phenotypic space as an emergent property of biological regulatory networks
Abstract: Forthcoming
Theory Lunch w/Zev Gartner "Order from disorder: self-organization in development and disease"
Zev Gartner | University of California at San Francisco
Title: Order from disorder: self-organization in development and disease
Abstract: Forthcoming
Department Seminar: Dan Rokshar
Professor, Department of Molecular and Cellular Biology, University of California, Berkeley
Talk Title: "The past has left its traces on the world.”: deeply conserved synteny and the evolution of animals
Host: Marcos Simoes-Costa
Food to be Provided
Pizza Talks: Rikki Garner
On Zoom and in WAB 563
Pizza Talks: Rikki Garner, Megason Lab
Talk Title: Cell migration: Noise and robustness in spatial patterning across scales
Theory Lunch w/Matthew Fisher "Quantum processing in the brain"
Matthew Fisher | Kavli Institute for Theoretical Physics UC Santa Barbara
Title: Quantum processing in the brain?
Abstract: The endeavor to construct a laboratory quantum computer has evolved into a multi-billion-dollar undertaking. However, an intriguing question arises: could we ourselves be quantum computers? While maintaining quantum coherence on macroscopic time scales is exceedingly unlikely in the warm wet brain, there is one exception: nuclear spins. Our strategy is one of reverse engineering, seeking to pinpoint the biochemical substrate and mechanisms that might underpin the hypothetical quantum processing involving nuclear spins. Seemingly, a specific neural qubit and a unique collection of ions, molecules and organelles can be identified. I will present an overview of our (largely experimental) ongoing efforts to delve into this captivating realm of exploration.
Department Seminar: Sahand Rahi
Assistant Professor, Laboratory of the Physics of Biological Systems, Institute of Physics, EPFL
Talk Title: “Why and how the DNA damage checkpoint gambles with cells' future + Preview: evolving switchable proteins.”
Host: Galit Lahav
Food to be Provided
Theory Lunch w/ Guillaume Montagnac "Associative memory in human cancer cells"
Guillaume Montagnac |Gustave Roussy Cancer Campus |Paris, FRANCE
Title: Associative memory in human cancer cells
Abstract : In psychology and neuroscience, associative memory refers to the capacity to learn and remember a link between two unrelated items. Although associative memory is widely believed to be restricted to animals possessing a complex nervous system, several reports have suggested that single-cell organisms can be conditioned to develop an associative memory-like behavior. Here, we report that human cancer cell lines can be conditioned to associate an extracellular matrix component and Gefitinib, a drug that reduces cell migration velocity. Collagen-I was periodically paired with Gefitinib and we observed that conditioned cells progressive decreased migration velocity on collagen-I but not on other extracellular matrix components. We identified the adenosine receptor ADORA2A as a key actor regulating the acquisition of associative memory. We also observed that the magnitude of the conditioned response oscillated over time with the same periodicity as paired stimuli presentations during conditioning. We found that mitochondria morphology oscillated with the same periodicity, suggesting that memory and energy metabolisms are linked. We propose that human cancer cells can be conditioned to integrate a link between two stimuli from their environment in a process that may allow to anticipate future stress exposition.
Pizza Talks: Michael Cory
On Zoom and in WAB 563
Pizza Talks: Michael Cory, Silver Lab
Talk Title: Rare No Longer: Harnessing the Power of Biology for the Sustainable Purification of Rare Earth Metals
Theory Lunch w/ Gautam Reddy "Dynamic landscapes during cellular growth and diversification"
Gautam Reddy | Department of Physics | Princeton University
Talk Title: Dynamic landscapes during cellular growth and diversification
Abstract: The complexity of gene regulatory networks in multicellular organisms makes interpretable low-dimensional models highly desirable. An attractive geometric picture, attributed to Waddington, visualizes the differentiation of a cell into diverse functional types as gradient flow on a dynamic potential landscape, but it is unclear under what biological constraints this metaphor is mathematically precise. In this talk, I will show that gene regulatory strategies that guide the growth and development of a single cell to a target distribution of cell types are described by time-dependent potential landscapes, under certain specific growth-control tradeoffs. The theory highlights a conceptual link between nonequilibrium thermodynamics and cellular decision-making during development.
Theory Lunch Zoom Details:
https://harvard.zoom.us/j/94761157754?pwd=kvdkMldvwFaZAeMOvXzMPeP5Pyne30.1 | PW: 10042024