Property:Post-meeting notes
From Complex Time
This is a property of type Text.
I
# Presentation highlights: (i) The role of constraints in phenotypic evolution as means of generating irreversible evolutionary endpoints and set upper limits to evolutionary trajectories. (ii) Role of constraints in species' ability to adapt to changing environments. (iii) Species come up against hard limits to phenotypic plasticity under climate warming. (iv) In order for thermal reaction norms to evolve in the face of climate warming, there has to be genetic variation. Unclear that reaction norms under strong biochemical control (e.g., development) have sufficient amounts of variation for the upper thermal limit to evolve in response to warming.
2. Open questions:
2.1 Connection between Darwinian adaptationist evolution and the idea of increase in disorder (as in the second law of thermodynamics)
2.2 What exactly are irreversible evolutionary endpoints? Can we come up with a specific definition of irreversibility?
2.3 Selection and constraints are not the same thing. This needs to be clarified.
3. How my perspective has changed: I want to think more carefully and deeply about the connection between Darwinian evolution and the second law of thermodynamics.
4. Reflections on other presentations
4.1 Stephen Proulx - I very much liked this presentation about the population genetics of low-probability transitions. I was particularly interested in stochastic selection due to lottery competition that leads to alternative stable states making it possible for mutations of large effect to cause transitions between states in a directional manner. I also liked the models of stochastic tunneling or valley crossing, that provide possible avenues for transitions between states. The case of multiple independent mutations enabling valley crossing is equally fascinating. I particularly liked how the examples shown related to the central theme of irreversibility and transitions.
4.2 Dervis Can Vural - An elegant presentation of the evolution of cooperation against the backdrop of fluid dynamics. I would like the theory to be generalized to perturbations other than shear so that it can also apply to pathogenic microbes within a host and other situations that do not involve fluid as a medium. I think you also should take the plunge and try to connect this theory to Hamilton's theory of kin selection. It is hard, and perhaps not analytically tractable, but it would be worth doing.
4.3. Samraat Pawar - I like the connection between metabolic constraints on species interactions and carbon fluxes.
4.4 Fernanda Valdovinos - The idea that adaptive foraging by mutualists (e.g., pollinators) allowing the persistence of nested mutualistic networks is a novel and exciting finding that pushes the field forward.
'''Annette Ostling:'''
Suppose that more similar species compete more directly - a very reasonable assumption given the reality of the fuzziness of defining species. Then an LV equation can have species parameterized by "traits." Different "niches" arise naturally in this model. A study of traits in a Panamanian forest suggests that species cluster in trait niches. ''Question'': the niche peaks tend to "repel" each other - is it worthwhile to think of individual peaks as species "quasiparticles" that interact with nearby quasiparticles through some generalized interaction, thereby giving rise to large scale, slow time dynamics?
'''Otto Cordero:'''
By creating nutrient beads and immersing them in seawater, colonization by bacteria can be studied in a controlled way. Genomic signatures can be mapped to particular strategies - degraders, cross-feeders, and cheaters. ''Question'': Does the geometry of the beads have an effect on the time behavior of populations? In other words, if instead of spheres, the bacteria were left to colonize tori or sheets, would there be any noticeable differences?
'''Priyanga Amaraserkare'''
Interested in when species are able to adapt to new environments. One take away is that species die much more quickly when introduced to high temperatures as compared to low temperatures, and that rate controlled processes have a different functional dependence than regulatory (as a function of temperature). Another very interesting point: it is very common for species from the tropics to invade more temperate climates, but it is much more unlikely for a temperate species from higher altitudes to invade a tropical environment. +
'''Darwin's arrow of time versus the 2nd law'''
Life on earth is subject to constant energy input from the sun. The 2nd law of thermodynamics, that entropy should increase, is for a closed system. So it seems it is not even really relevant for thinking about life on earth.
'''Definitions of irreversibility'''
One key definition of irreversibility we discussed is that if you reversed time the process would look strange—abiological. Can we make that definition more quantitative? Perhaps we mean simply that it would be going against changes predicted by the selective regime or expected population dynamics? Can we make that definition accommodate degrees of irreversibility, e.g. looking backwards involves changes less likely to happen? This fits in with what Priyanga talked about with adaptation to colder being easier than adaptation to warmer due to the shape of the relationship between the performance and temperature.
''BUT is this definition too broad?'' Any system with an equilibrium point is irreversible in this sense, because if you reversed a time series of it approaching its equilibrium it would not make physical/biological sense?
So do we really need to add something more to that definition, perhaps to include the idea that some environmental variable is being changed in time and we are watching the response to it, and asking if the system would go back if we changed the environment back? In that case our definition of irreversibility is the presence of hysteresis?
Is another definition of irreversibility that the system changes in a way that impacts its future potential changes or response to change in the environment? Or maybe this is just something often associated with irreversibility, as it is not the same as asking about a reversal of time, but instead whether there is path dependency in the system? Is this question of path dependency related to Gould’s question about whether replaying the tape of life would lead to the same outcome?
An example of this idea of the change in the system impacting potential response to future change is the case of competitive cluster formation (see below). If one assembles the community under one environmental filter, and then the environmental filter changes, community biomass may go down and never achieve what it was before or could have been under new environmental regime if assembled that way in the first place. The idea is that the change in environmental filter may not have be strong enough to overcome competitive footholds species have in the community.
'''Drift and selection, and is drift reversible?'''
Two key processes in ecology and evolution are drift and selection (among species or among alleles). Is drift in a sense a force creating more disorder? If so, we would think it would increase entropy in a sense and lead to irreversible changes? (Note there has been one paper by Sella and Hirsh in 2005 in PNAS trying to think about drift as something increasing entropy and more broadly a "free fitness" function like a "free energy" function, summarizing the role of selection and drift in the state of the population.) But we discussed it yesterday as reversible. Can we be more quantitative about why we think about it as reversible?
''Further, is drift really reversible?'' Drift can prevent a system from reaching another fitness peak, by causing loss of advantageous, neutral, or disadvantageous alleles when they are rare. (Recall the stochastic tunnelling examples Stephen Proulx talked about for how evolution may overcome this however.) So it can change the future possibilities for the system. It can also be involved in the somewhat irreversible process of competitive cluster formation (see below). A particular species may gain high abundance by chance (drift) and then have a stronger influence on the the competitive landscape for other species and become abundant in its cluster. Actually if there are no edges and no environmental filter, drift one of the two key ingredients in cluster formation (the other being initial conditions).
'''Key idea related to irreversibility and questions I raised in my talk'''
In my talk I highlighted that the formation of species clusters on trait axes under competition has some degree of irreversibility, in the sense that under strong competitive sorting, once a species dominates a particular cluster it is unlikely to loose its foothold. It would take a strong perturbation in species' abundances, or a change in which species are favored by the environment, to change which species would dominate in each cluster. Further, once certain species have gained a foothold in each cluster, this influence any subsequent assembly or evolution (selection, speciation, extinction) in the community.
The questions I posed about this particular phenomenon of irreversibility are:
1) How is the rate of competitive sorting, i.e. the strength of cluster formation, and hence degree of irreversibility, shaped by the mechanisms of competition? Do clusters emerge for all realistic competition mechanisms?
2) How will cluster formation depend on spatial scale, and how will this be influenced by the strength and scale of dispersal, relative to the scale of any heterogeneity involved in niche differentiation mechanisms?
3) Is the irreversibility of community pattern formation a particular concern for communities that may become isolated? These communities will experience extinction debt, and afterwards their resilience to environmental change may be low (the species that may be favored by the new environment may be gone).
H
'''Day 1'''
<u>DPromislow:</u> Evolution shapes function and failure. Three dimensional space: Failure, Function and Evolution.
Main question: why do different agents age at different rates (faster, slower?)
<u>MHochberg:</u> Function criticality, aging and resilience. Coupling mechanisms of adaptation and aging. Wait, what are we referring to here for adaptation?
<u>James DeGregori</u>
Is it evolutionary Explain this! Oncogenic mutations in young healthy stem cell
populations typically reduce cellular fitness.
Peto's paradox: large and long-lived animals do not develop more
cancers.
Cancers requiring different numbers of driver mutations and
originating from vastly differently organized stem cell pools
demonstrate very similar age-dependent incidence.
Adaptive oncogenesis: we evolved stem cells that are well adapted to
the tissue niche. Stabilised selection for the evolved type. This is
more powerful than avoiding mutations. Stem cells would be adapting to
new environment.
q: does it mean that changing the tissue environment you would lead
stem cells (cancer) to evolved towards optimality in the "healthy"
environment?
q2: how about metastasis? do they evolve new-niche specific
variations? or physiological adaptations?
<u>Rozalyne Anderson</u>
Conserved pathways responsive to caloric restriction, across tissues.
Chronic low-level mitochondrial activation in cells. Mito regulator
PGC 1A (expressed to 1.5X, similarly to CR tissues). Super cool!
<u>Tulja:</u>
#Frailty and disability: transitions
#Post-reproduction lifespan
u(a, H) < u(a, S)
u: death rate
a: age
H: health
S: Sick
Recovery becomes harder as we age.
Decline in homeostasis with aging.
Potential well model (similar to fitness lanscapes)
Escape from potential well. Fluctuations depend on well depth, curvature of well and size of fluctuation. It's a multidimensional space.
Prob (H --> Ha)
Prob(S --> Sa)
As we age, the profile of the H and S well changes, making it harder for the ball to escape from the S well.
Longitudinal data on health, frailty, morbidity. HMM drive transition states. (Aging studies, Framingham, HRS, ...).
borrowing Hamilton's fitness (grandmother effect, old men, learning, transfers).
<u>Sabrina Spencer</u>
Causes and consequences of non-genetic heterogeneity.
How do cells switch back and forth between quiescence and
proliferation. Shift from G1 towards G0 depends on CDK2 (off). She develops awesome trackers for CDK2. Wow!
Bifurcation in CDK2 activity appears in many cell types.
p21 inhibits CDK2. p21 -\- remove the bifurcation, leading to proliferation
only, no quiescence. Many quiescent cells have DNA lesions. Mothers of
quiescent daughters have a longer cell cycle. Daughter quiescence is
decided in the mother's G2 phase. Mother cells pass DNA damage to
their offspring cells (i.e. do not retain the damage).
CDK2low state fortifies cell lineage against stress.
<u>Barbara Natterson-Horowitz</u>
Cardiovascular disease in humans, non-human primates, lions sympathetic and parasympathetic system.
Sudden cardiac death (SCD), "Tahatsubo" cardiomyopathy.
High adrenergic events. Sympathetic responses are flight and fight. Parasympathetic responses are opposite: faint, shit yourself, etc.
Fainting: vasovagal syncope (VVS), due to underprofusion of the brain.
VVS: paradoxical bradycardia has developmental characteristics: depends by the age of first fainting.
Alarm bradycardia is preferentially a juvenile phenotype.
Early life events: shaping autonomic nervous system for later events.
'''Day 2'''
**Intro by Michael Hochberg
- Basins of attraction - Failure, gradual, rapid, can we predict it - What is failure, underperformance or misperformance - Tradeoffs and pleiotropy - Networks - not discrete comparts - Complexity vs. simplicity - Henry Ford Aging and senescence, time and disfunction. SSpencer: Immortality, regeneration and rejuvenation. DPromislow: Michael Rose's concept of age-depedent decline in fitness components.
**Morgan Levine
Geroscience. Hallmarks of aging that accumulate over time. Systems biology of aging, changes ion the moleucular level propagate through the networks. Preservation of function (paper with Luigi Ferrucci in Circulation research). Aging heterogeneity, chronological age is an imperfect proxy of the latent concept, biological aging. Epigenetic clocks. Chronological age has been shown correspond with distinct chagnes in DNA metylation at specific CpG sites. Horvath's and Hannum's clocks. These clocks are done using supervised machine learning. She wants to capture the true residuals, rather than minimizing the residuals. Understand whether those residuals tell you anything about the underlying biology. Getting more biology than just chronological age from the clocks. Phenotypic age, predictor of aging-related mortality based on clinical measures. The used markers are albumin, creatinine, glucose, c-reactive proteins, lymphocyte percent, mean cell volume, red cell ditribution width, alkaline phosphatase, white blood cell count, age. Consensus co-methylation networks - data from 6 different tissues. 14 modules, showing age correlation for each CpG. Which piece of the clock matches what tissue, etc. Diseaseome map.
**David Schneider (Stanford)
Resilience and disease space. How sick are you going to get with a given load of pathogens? Symptoms = f(microbe load). It's a disease tolerance curve. However, tolerance is a population measure and does not apply to individuals. How do we expect disease dynamics to vary? All models are wrong, but some models are useful (George Box). [[https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1002436][Tracking Resilience to Infection by Mapping Disease Space]]. Using a range of mouse diversity to identify imporant regulatory mechanisms. 8 parental mice used to generatel RIL. He brings up that methylation clocks are actually calendars, not clocks.
**Marten Scheffer
How ecologists have been looking at resilience. Resilience: the capacity to get your acts together, to recover after damage. How quickly things come back to normal after perturbation. Sometimes things do not recover from perturbation and does not come back. Another way to think about it is that the system has become brittle. Unstable equilibrium. Theory developed in the 1960s, catastrophe theory. Salvador Dali and his own tipping point. Holling in 1973 thought about resilience. How to know a stability landscape. Tipping elements in the human intestinal ecosystem. Generic early warning signals = indicators of resilience. Critical slowing down. Dinamic indicators of resilience (DIOR). A bird shits on your head. After a hour you're ok, it means you're resilient. If after a hour you still hungry, you're not resilient, you may be depressed. This is based on patterns of micro-recovery. Fast recovery, slow recovery. Temporal autocorrelation help you predict tipping points. The previous arrow of time group has looked more at the complexity. This group is more looking at the molecular side.
**DPromislow
Common dynamical phenomena underlying description of the causes of failure. What are those communalities? This can help us get to the causes and dynamics of failure. This could be an argument of discussion. Group discussions to address communalities and differences in failure.
* Group Discussion 1
Can we reverse, halt, or slow down aging or is aging inevitable
- which components? Epidermis.
- ways to slow aging: CR
Is aging inevitable? Entropy. Does entropy at individual level
increase, causing aging? Individuals are open systems.
What are the components of aging?
How do we distinguish between intrinsic and extrinsic causes of aging?
Intrinsic: failure in DNA repair, proteostasis.
Extrinsic: falling down, UV light
Do you really return to the same initial state after a perturbation,
or are you missing a dimension.
The landscape that David Schneider showed may change as aging.
progresses. Contours over which these things occur may be morphing.
'''Day 3'''
C
'''Impact of the meeting on my research'''
The main thing I took away from the meeting was that robustness to damage is property of neural networks that is important from a theoretical as well as an applied perspective. I'm motivated now to think about revisiting the old method of lesioning and perturbing with noise neural network models, so as to better understand the degree to which they are sensitive to local damage and to displacement of their dynamic trajectories.
I would also like to explore how robustness to damage and noise relates to robustness to adversarial attacks and to robust generalization performance in novel situations (i.e. changes to stimuli or more generally the behavior of the environment). Before this meeting, I would have conceptualized these different forms of robustness as largely unrelated. I would have thought that the commonality suggested by the fact that they can all be considered forms of robustness is largely misleading.
Now I think there may be deep links between robustness to damage, internal noise, changes of the body, and changes of the stimuli and environmental behavior.
Some half-formed hypotheses:
* Neural noise may help a neural network learn solutions that are robust in a variety of ways.
* Similarly, changes of environmental behavior (including stimulus statistics) during learning may help a neural network learn solutions that are robust in multiple ways.
* Predictive completion of partial neural representations (in symmetric, i.e. energy-based, or asymmetric networks) may provide robustness through redundant representation as well as enabling unsupervised learning through selfsupervision.
I'm quite keen to explore these ideas further.
'''Important open questions'''
• How can states close to criticality serve computation in neural networks?
• Might robustness to damage result from the same mechanism that enables robust generalization to new domains? (And what is that mechanism?)
'''Notes on the meeting'''
The discussion, though inspiring, was a little more wide-ranging than is optimal for making concrete progress.
It might good in future meetings to focus, and to more clearly and specifically define the topics of sessions and particular presentations.
P
'''Most useful thing I've learned''': The idea of outlining several causal models and the predictions that they can make about a certain phenomenon of interest, to then test which one fits the data better. This type of deductive reasoning is, I'd say, underused in my field. I'd like to think again about the causal models for fertility and to review potential causal models for epidemiologic transitions. However, at the same time, the conclusion from the lecture (and the paper) on these causal models for fertility ended up being about the importance of considering all models simultaneously, or, even better, considering how they have a differential effect over time. That is, some of the models may have an importance in the beginning of the fertility decline while some others may be more important in the later stages. I think we need to leverage the idea of feedback loops and endogeneity in complex systems to better accommodate the presence of multiple causal models.
'''Planning to use knowledge''': I plan to use the framework on migration determinants, especially at the macro level, and to leverage some of the data sources that were mentioned in the migration lecture. In particular, I'd like to see the connections between internal migration (rural to urban) in the determination of urban health outcomes. I'd also like to incorporate some of the lessons about causal models, but applying them to epidemioloic transitions.
'''Interesting conversation''':
* On the necessity to consider endogeneity in the causes of population dynamics. X affects Y than in turn affects X again.
* On the necessity to consider inequalities and distributional effects +
'''Most useful thing learned in course:'''
For me, the most useful about this course was seeing more examples of and talking through the process of moving from 1) establishing research questions to 2) identifying the key variables to 3) developing the equations to relate the variables to each other to 4) developing models of how the systems work. For me, steps 3 and 4 have always been the most difficult and are currently the steps I am thinking through for my dissertation. Getting to talk with other researchers about this process and seeing how they developed equations and models to capture and depict their topics of interest has helped me make some inroads to get started on this process in my own research.
'''Some additional useful things learned in the course:'''
-The importance of always being aware of and periodically re-assessing/re-identifying the interplay between 1) models/abstractions of data patterns and 2) on the ground interviews and data collection that reveal the key variables/driving factors and logics/frameworks of the subjects that contribute to the observed data patterns (such as in Caroline Bledsoe’s . Without this ground-truthing and finding out what variables actually matter for the study organisms and how they conceptualize them, our models will be flawed due to either missing key variables or not being able to actually explain the implications of the patterns they show.
-The importance of considering time lengths/durations in a process. For example, as we saw in Mary Shenk’s lecture on demographic transitions, the same pattern (e.g. declining fertility rate), but over different time scales (e.g. a longer time in one area than another), can have significantly different impacts (e.g. amount of population in each of the two different areas). Another example from this workshop was thinking about what is the temporal resolution of our data and how this aligns with the temporal resolution of the variable of interest (such as Lori Hunter discussed in terms of census data and temporary migrations).
-What models to use for particular demographic problems (and how to accommodate variation and additional parameters)
'''Applications for my dissertation research:'''
As I’m building models and thinking of variables to consider for my dissertation, which deals with agricultural adaptations in light of socioeconomic changes in the indigenous American Southwest, this workshop will be extremely helpful (particularly Chris Kempes’s work on the land/resources needed to sustain a given population and Charlotte Lee’s model integrating environment, population, and society).
I
'''Pamela Martinez'''
Connection between population dynamics and strain diversity. What is the response to pathogen intervention? (clear in a Chessonian framework, unclear when there is strong strain diversity). 1) Rotavirus. Robust antigenic diversity. Estimate of the parameters from cases: G-type strong specific immune response, P-type strong general immune response (concern: very few data, enough statistical power). Surprising: no effect of vaccine/intervention. 2) S. pneumonia. Frequency of strains. How does frequency change after intervention? Replicator model, predicted fitness from genes (how? Not clear): can predict frequencies after intervation (which again does not work). Questions: 1) what determines the equilibrium frequencies of genes? Why loci are under negative frequency negative dependent selection?
'''Annette Ostling'''
Stochastic open competitive communities. Niche/unique opportunities/deterministic forces vs chance/neutral/stochastic forces. Finding niche differences is a challenge. Stochastic Lotka-Volterra on 1dim niche axis . Cluster emerge. 1) Metric. What is a good measure of clustering and how is that affected by different mechanisms? 2) Data. BCI and regional pool.
'''Greg Dwyer'''
Gipsy moth & virus/fungi. Inference of parameters from time-series. Key ingredient: variability of susceptibility between individuals. Works in reproducing data (but it is hard to predict)
'''Otto Cordero'''
Hyper-diversity of microbial communities. How do we interpret that diversity? Patches of organic matters (~detritus) are hotspots of diversity: multiple species are recruited. Lab experiment: biopolymers, 100 species. -Omics + culturing + phenotyping in order to recostruct dynamics. Observed successions (are true successions? Can also be explained by variability in growth/dilution rate). How succession depends on niche-breadth? Idea: niche breadth is bimodal (small niche breadth: early colonizers, specialized. Large niche breadth: late colonizers, cannot be grown in culture). He then looks at the metabolic networks: close to the central metabolism everything is more homogeneous, close to the periphery everything is more heterogenous (high variability in copy number). He then plots #of chitinases vs the # of coevolved modules. Early colonizers are high in both. Late colonizers are low in both. Free loaders have high # of coevolved modules but low #chitinases (can eat what free riders eat). Ecological dyanmics is consistent with this classification (degraders grow initially and then decay, cheaters and cross-feeders grow later)
'''Robert Marsland'''
Two facts: A) Microbiome is very diverse. B) Taxa is unstable, function is stable (robust patterns). Assumptions: 1) classical ecological models are inadequate for understanding ecosystems 2) large / diverse ecosystems are typically random. (Random) Chemostat model predicts many patterns. Why random works well? Above stability threshold random works very well.
'''Priyanga Amarasekare'''
1) Phenotypic traits are the interface between organisms and the environment 2) Evolution: (mutation+)contraints+selection 3) Irreversible processes arise from contraints. Different types of contraints: genetic (evolution acts only on heritable traits), energetic (tradeoffs), morphological (upper limits to evolutionary trajectories). Rest of the talk on how different traits depend on environment (Temperature). Two mechanism for reaction (to T variaton) norm: enzyme activation (monotonic response, e.g. mortality) and regulatory/hormonal (unimodal shape, e.g. eggs maturation) -> conserved across taxa.
'''Fernanda Valdovinos'''
'''Samraat Pawar'''
'''Stephen Proulx'''
'''Dervis Can Vural'''
'''OPEN DISCUSSION'''. 3 themes/questions
[[File:OpenDiscussion.jpg|center|450px]]
<u>1) Identify quantities that are always increading or decreasing.</u>
(side questions: why? At what scale (temporal, spatial, ...) do they have that trend?
Possible examples (in a Markov chain): return time (as a measure of irreversibility), turnover time, how often the system returns to the original state, fraction of trajectories that go from A->B vs from B->A.
Possible examples (interpretable quantities): # of species, biomass, # of limiting resources, # of niches, properties of resource consumption (e.g. efficiency), degree of specialization, interdependence, major (evolutionary) transitions
[[File:Question1- quantities to measure.jpg|center|450px]]
<u>2) List of properties that we expect to be reversible/irreversible</u>
Irreversible: oxygenation event (major transition), increase of specialization, latitudinal gradients, adaptive radiation, cluster formation
Debated: drift, mass exctintion (in what sense they are irreversible)
Reversible: function(?), total biomass (?)
[[File:Question2b.jpg|center|450px]]
<u>3) (relevance of) transients</u>
It depends on the timescale of evolution vs population dynamics.
Relevance of timescales: timescale of evolution, environmental change, behavior vs timescale of population/community dynamics
Relevant examples: patterns of extintions (position in transient determines outcomes), cycles ("always" in a transient)
[[File:Question2.jpg|center|450px]]
'''Pamela Martinez:'''
Very interesting research on how strain diversity can affect disease spread. I learned a lot about best practices on how feetting models to data because of the discussion this work provoked.
'''Annette Ostling:'''
Impressive work of testing the prediction of a competitive model of plant species with empirical data. The authors found that clustering happens in tropical forests due to niche partitioning. I learned much on different competitive models.
'''Greg Dwyer:'''
This talk was very useful for me to understand ways in which empirical data and models can interplay to make concrete predictions that can inform management. The applied case of informing agencies when to spray the forest with virus to stop the tree disease was very illuminating.
'''Otto Cordero:'''
Interesting application of ecological theory to microbial communities. I really enjoyed the way the speaker identified biological mechanisms in his empirical system and was able to connect the modeled dynamics to those empirically tested mechanisms.
'''Priyanga Amarasekare:'''
This talk made me think in a deeper way about constraints on phenotipic/genotipic variation that can help us understand how ecological system may respond to human perturbations such as climate change.
'''Fernanda Valdovinos (my talk):'''
It was extremely helpful for my research the in depth discussion that the audience provoked on the details of my model. The dissecting questions I received on my equations and their consequences were very illuminating. I will definitely use some of the new understanding I acquired trough answering those questions in the paper I'm currently working on. I also really appreciate the philosophical question that Greg asked me over the break and Jacopo helped to answer. That question was about what are we actually learning by using a network approach instead of just many differential equations as we have been doing for years in ecology prior networks. I would really like to further discuss this question as a group tomorrow.
'''Stephen Proulx:'''
Amazing talk that helped me better understand adaptive dynamics, how we can read mutation/invasion maps and how to make better use/understanding of fitness landscapes. It was fascinating to me the trade-off example on plant fertility-survival that showed a clear case in which small vs large mutations can drive the genotype/phenotype of plants to different attractors. I also liked a lot one of the speakers questions on how to produce general theory from non-equilibrium cases and the discussion that question provoked.
'''Questions from my talk:'''
True or false?
1. Classical ecological models are inadequate for understanding microbial ecosystems.
2. The large-scale, reproducible patterns we see across microbiomes are emergent features of “typical random ecosystems.”
3. Diverse communities will almost always behave like “random ecosystems.”
'''Question I want to discuss'''
Can large-scale ecological changes over time be understood through simple general principles? (Samraat, Jacopo, Priyanga, me: flux balance/metabolic rates, random matrix theory, biochemical and morphological constraints)
Are there ecological summary statistics that change monotonically over time? (Jacopo, and some of my work that I didn't talk about -- see references)
Which properties of individual organisms are essential for predicting large-scale ecological changes?
- Pamela -- why do some vaccines fail to produce large-scale ecological change (i.e., pathogen extinction)?
- Greg -- importance of distribution of susceptibilities to dynamics of epizootic onset.
- Fernanda -- importance of adaptive foraging for understanding why pollinator systems don't collapse -- and why they might under new circumstances.
- Samraat -- taking the mechanics of inter-species interaction seriously, understanding how they affect response of ecosystem to temperature changes.
- Priyanga -- asymmetry of reaction norms for temperature variation.
What do we miss when we focus too much on equilibrium/steady states? (Prevalence of cyclic disturbance/recovery dynamics -- Greg, Samraat) +
W
'''SFI WG Big Unanswered Sleep Questions'''
1. What is sleep? Behavioral definition? Phenomenological/experiential definition? EEG/physiological definition? Which metrics—fitbit, online, over what time span, Etc.? How closely matched in terms of time of onset? How closely matched in terms of functional or health effects? How do we separate sleep ability from sleep need? Do we need a species-specific definition and can we come up with a very general definition that abstractly covers everything? How strict should we be with definition of REM and SWS by voltage or time-scale and frequency characteristics across ages versus qualitatively similar behaviors? What are the building blocks of sleep? When do definitions fail? What breaks these definitions? Which questions correspond to which definitions? Can we make that mapping? Is there a universal rescaling of EEG or something else where sleep does look the same across species, life stages, etc?
2. What are the functions of sleep? Can we construct a phylogenetic tree of life and developmental stage tree of life for the function and features of sleep? Are all about evolutionary/energy optimization of processes that don’t happen as well during waking? Restoring functions vs learning functions vs other
a. Feel rested/cure sleepiness
b. Performance of tasks
c. Emotional regulation
d. Metabolic function
e. Maintenance of brain and/or body
f. Immobility during dangerous times for predation etc.
g. Memory and information processing
h. Energy conservation, adaptive inactivity
i. Immune
j. Inflammation (molecular) and pain (behavior/perception)
k. Endocrine function
l. Growth (SWS)
m. Protein homeostasis
n. Endogenous sensory input with higher signal/noise than otherwise for neonates
o. Maintain mitochondrial function
p. CSF/glymphatics clearance
q. Synaptic homeostasis
r. Breathing and cardiovascular regulation
s. Sleep cycling for vigilance to monitor fires, danger, etc.
t. Appetite control
u. Increased/decreased predation
v. Allergies
w. Kidney function
x. To dream
Consequences of sleep. These would be costs/tradeoffs in an optimization function
a. Reduced alertness
b. Reduced foraging
c. Reduced social contacts
d. Reduced activity
e. Fasting
f. Reduced information gathering
3. Does evolutionary origin for function of sleep correspond to current functions of sleep? Is one sleep function dominant and the rest subservient or piggybacking? Are all on equal footing now? How big is the selection pressure in terms of evolutionary theory and measurement? Is there a group-level selection for sleep? What is first common ancestor of sleep? Circadian rhythm came before “sleep”.
4. Do different functions compete for time? Which function has longest time constant, which is distinct from which is most important? How does this change with life stage?
5. Are there different amounts of sleep time required for different functions of sleep? IF sleep is limited, are these different functions competing for the available time? If you restrict sleep, what is the ordering and magnitude of loss of different functions? Much known for learning and memory but what about other functions?
6. What are the characteristic time scales of sleep and what sets them? Which ones are invariant, and which ones change across size, species, developmental stage, brain region, task, etc.? What data already exist for this? What were our models predict? If you take differential equations models and parameterize by correct time constants (based on scaling of neuronal scaling and axon conduction velocities or based on physiological/allometric scaling or both), as you tune them what dynamical shifts or phase transitions do you see? Why does a 90 minute nap help more with learning tasks than 5 hours of sleep? Go from fastest (ripples) all the way up to long timescales. Can we list characteristics timescales for each function above?
7. Do we mean mechanism or function? How are these all reconciled? All equivalent?
8. If sleep is for repair, does whole body repair work differently or over different time scale than just for brain? Immune response, wound healing, and bone density connection to sleep and circadian rhythm? How to quantify?
9. Does brain “saturation” cause sleep? Global saturation (no?) or local saturation? How are local limits set and what sets partition of space for each category, say physics versus French literature? Seems like it should be experimentally measurable/testable now. What exactly needs to be measured? Why don’t we do it?
10. Do we need and can we get better developmental sleep data?
11. How do we include effects of temperature in sleep models? Via metabolism? Via loss of thermoregulation in sleep, hibernation, torpor in birds? What about sleep changes in total time, %REM, sleep cycle, etc across body temperatures in ectotherms, even for similarly sized organisms.
12. Is dreaming a way for brain to devise better computational algorithms for exploring large dimensional data or parameter space for theory? Why does this require sleep? Is it because it isn’t accessible to our conscious brain because done by some other part of the brain and takes lots of energy so have to save energy from elsewhere? If it’s so useful, why do we forget dreams? Maybe because their actual content isn’t useful? Maybe if you remember too well, you start to confuse dreams with reality and that becomes big problem? Is there some wake correlate or default state or meditation or exercise in which something similar can occur?
13. What are the individual and Societal health implications of all the above?
14. Does sleep time change with city/community/social network/college size?
15. How can we best leverage large digital datasets? Can we align 23andme with fitbit and other data?
16. Can we develop a quantitative, mathematical, predictive, universal theory of all this? What would it take to do this in terms of data, assumptions, mathematical and computational tools?
17. How can we capture the heterogeneity of the data? Deviations from universal or described on its own terms or correlates and cross-effect for different features. Can we isolate which factors impact sleep? Gender, genome, SES, race/ethnicitiy, age, etc. Higher-order interaction effects.
18. How do recent experimental tools enable the answering of the questions above?
'''What/Who are we missing?'''
Ken Wright UC Boulder Metabolism and Microbiome
Eve Van Cauter U of Chicago Metabolism and Immune function
David Paydafar UT Austin Engineering Mathematical Biologist and modeling of SIDS
Rosemary Braun Northwestern Statistical Analysis of Circadian Phase from 1 or 2 samples
Phyllis Zee Northwestern Sleep in elderly (bug picture thinker/ great at networking)
Michael Perlis U Penn Insomnia and suicide
Daniel Buysse Sleep health
Sonia Ancoli-Israel UCSD sleep and chemo
Laura Lewis Harvard Clearance of CSF during sleep
Monica Haack Harvard Immunology/Inflammation/Sex differences
Mark Opp UC Boulder
More traditional Evolutionary Biologist
Book audience more towards industry, Gates foundation, funding agencies, scientists, CDC,
Present more scientific basis and evidence-based approach to importance of sleep and understanding it and its consequences.
Mostly about basic science questions we’ve discussed throughout but with bullets at end of each chapter about implications and a chapter or two at end that is more focused on implications for health and policy.
Make more focused in terms of modeling and timescales of sleep as theme of book.
Data generated at many timescales, timescales can be organizing principle but must reconcile and synthesize them, modeling needed to reconcile and synthesize these timescales, this process actually unique vantage point for answering and asking big questions about sleep.
Help bring model-resistant scientists to understand importance of modeling.
Let’s be provocative and say we can try to model dreaming!
How does a brain calculate at the neuronal level which memories are valuable?
How do we model sleep across timescales?
Connecting machine-learning to modeling approach to make it better
Importance of working and talking across disciplines
How do a mapping from questions to appropriate scales. Hard to do it as just individual group or lab.
'''Talk Notes:'''
'''Susan Sara:''' Locus Coerelus, PreFrontal Cortex, Sleep Spindles all increase about 2 hours after leaning task in rats. Higher LC firing rates even during Slow Wave Sleep, suggesting important learning and consolidation happening during sleep. LC is silent during REM sleep. Much more firing together during SWS than during wake. What sets time scales of 2 hours? It is task dependent because some take and hour. Does characteristic time scale change with species or brain size? What sets this time scale? Does it change across development? What is theoretical expectation? Any data? How to disentangle age effect from increasing difficulty of task being learned. Must be careful. Coordination and coupling of spindles and ripples. Which is first? How are coordinated? Which is functionally more important? Temporal relationships. Can replay happen without ripples? Replay in visual cortex is not related to ripples. Not clear ripples and reply are causative for memory. Maybe it’s the subsequent ripple-spindling coupling that’s disrupted instead of the present one.
'''Gina Poe:''' Norepinephrine at Beta Adinergic receptor is to increase/encourage long-term potentiation. Slow Wave Sleep goes aing with absence of Acetochioline even in unihemispheric sleep in seals. Define sleep as a behavioral state but maybe that’s hard because of dissociated states. Maybe sleep must be defined as functional question to ask what’s needed or what counts. Is millisecond sleep enough to do something functionally? Lack of norepinephrine is required to erase memory so erasure can only happen in sleep. And targeted erasure (meaning targeting memory) can only happen in REM sleep thru depotentiation. Takes about 5 days or learning mazes for consolidation and reversal of learning. Again what sets this time scale across task, species, development, or amount of brain involved? What tags or systematically decides what information you want to remember and what you want to forget? Firing rate lowest in REM sleep. Hard to see sleep in insects etc. because electric signals cancel out because not layered or striated or organized in way that gives clear signals.
Is familiarity versus episodic memory two separate states or is it a continuum? Familiarity might be the feedback to the hippocampus that this can be eraser. Seems like two separate states. Why do we want to forget? Storage limitation, not useful, scary/traumatic, updating of previous thought like Santa Claus, etc. Danger of seizure from overload or saturation but seems like little evidence for this. Maybe saturation is only at the local circuit level. Flips questions around to ask why do we partition brain memory in way we do and why we do give this chunk of memory to this specific task. Nobody has looked at this question experimentally but could be done now. Maybe not really forgetting anyway but more like downweighting certain memories compared with others. Very modular structure and organization to cholinergic and adinergic systems. Perhaps people with super memories have less redundancy of memory. Would they forget more in old age? Is this observed?
'''Sara Aton:''' Foot shock creates fear memory that isn’t consolidated if transcription/translation is disrupted, neuron firing is blocked, or sleep is not allowed. Increase in theta and ripples (so oscillations in general) after learning event across sleep. If you induce similar pace (7 Hz) oscillations you recover consolidation even without sleep. Not clear of side effects of what happens to other functions of sleep if this is all you do. Use of optogenetic tools to do cool experiments. Able to make it so there is fear response but without awareness of what they’re afraid of. Correct time scales of rhythms are main thing needed for appropriate spike timing and that happens during sleeps and that’s why sleep is needed for memory consolidation even if that’s not primary function (although it could be). Is the 7Hz timescale task dependence, species dependent, developmental dependent, etc? Going from rate code to phase code to renormalized rate code and strengthening connections to propagate information through circuit. Thinks 7Hz is circuit dependent and circuit frequency is tag in that way that depends on wiring. Like a resonance frequency. She thinks phase matters but not frequency but not clear. Could you decipher or reverse engineer these codes to read out dreams or other things that are happening. Seems like there are experimental tools where we could start to do that. Circuit itself and post-synaptic partners are themselves making decision of what to remember or forget. Is certain cells can’t keep they are in some sense filtered (probably literally) out? Could at least ask this question theoretical and see what happens. Does lag time correspond to spindles?
Slow oscillations are 0-1 Hz. Delta is 1-4 Hz. Slow oscillations go much deeper that allows for calcium rebound. Slow oscillations are more global and Delta are more local. Even individual cortical column or relay neuron can exhibit Delta by itself. Spindles are 10-15Hz layered on top of these other waves. Ripples are 80-200Hz on top of spindles. Gamma are 40-80 Hz and occur while awake. Only 15% of spindles coupled to oscillations even though learning relates to total number of spindles in sleep. Seems like paradox. 200mV for waves. Most power is in slow waves. How much of the brain’s metabolic rate does this take? A lot or a little? 1/f power happening for this. Measures of fractal dimensions of these time series? Do follow a power law. These are like temporal correlations. What about spatial correlations (using calcium imaging) or overlap in temporal and spatial correlations? What percentage neurons contribute to this? About 5% of hippocampal cells involved in a learning task and memory. 30% have place fields. About 20% decrease in glucose and oxygen consumption during sleep. About 37% decrease for humans in NREM. Would these timescales change across species. Similar in mice, rats, and humans. Calcium imaging could say more.
'''Kimberley Whitehead:''' Looks at spindle bursts that’s different from sleep spindles because also happen in wakefulness. People are now claiming to see spindles in adult. Maybe these two things aren’t that different and we just haven’t realized that yet. Help refine sensoricortical maps. If you prevent bursting, barrel cortex doesn’t get organized properly. Sleep-wake state on the scale of minutes for humans. Pre-term baby has loads of delta in all their sleep. Higher power in pre-term. Role of nREM becomes more important as you grow. Maybe phasic and tonic REM sleep have different roles. Sub-sleep state in scales of tens of seconds. Active sleep movement on scale of seconds. Sleep oscillations important for sensory cortical organization in gestation and early development. Log normal for frequency of wake bouts and active sleep bouts. Role of benefits of wakefulness versus benefits of sleep and which are more important for survival and selection. Sensory experience more intense during sleep than during wakefulness so reverse of sleep model for information input proposed in paper (Cao, Herman, Poe, West, Savage). Wakefulness set more by time to birth than time to gestation age. Why not plot versus body weight instead of age? Age explains more of the variance. It’s experience dependence response and processing which seems like the generic definition of learning, not just laying the substrate for it. Surge to breathe, arousal, etc at birth so that’s enough to switch dynamics of wakefulness and sleep.
'''Elizabeth Klerman:''' Sleep-wake cycle and circadian cycle Homeostatic cycle and circadian cycle must be coupled. Must desynchronize two clocks to decipher effects of each one. Homeostatic is need for sleep overall and not just circadian. Clocks shift for 25 or 26 hour timing. Scored in 30 second chunks of sleep or wakefulness. Organized around core body temperature minimum that typically occurs a couple of hours before you wake up. No circadian rhythm in SWS but there is in REM sleep. Not understood physiologically why that is? Tononi’s model is that SWS reverse buildup of LTP. SWS percentage is same regardless of sleep debt so total SWS decreases. Related I think to questions about %REM and %NREM across development and across species. Remember to ask this question during my own talk. Drive to wake strongest right before you go to sleep and drive to sleep is strongest right before you wake up. Like hanging on by your fingertips but need some force you’re fighting against then. Needed to help you consolidate sleep is hypothesis. Humans can know time of day even as seasons change by adjusting melatonin levels up and down. People used to be seasonally reproductive. Children and adolescents have much faster buildup of sleep pressure than older people. People who are out in bright sunlight during day aren’t as affected by devices at night. So something to do with change in magnitude of light, but just current amount of light. Different models needed for short exposure to light and effects on phase delay. Additional process coupled to process L or some type of modification of process L. How do go from average- or group-level models to more individual-specific models? How does caffeine or other drugs affect it? How does aging, disease, mental disorders, etc. affect it? What is physiological analogue of the two variables for the phenomenological model? Some amalgam of the physiological factors? Maybe different amounts of sleep are needed for different functions of sleep? Maybe they are competing ofr the time they get?
'''Cecilia Diniz Behn''': Sleep depends nonlinearly on amount of sleep deprivation and timing in circadian clock. Characteristic time scales affect shape of circadian waveform that affects timing of sleep and wake. Can we get a more physiological perspective and include that in modeling framework. Network between SCN, Wake, REM, and NREM. Choose scale at which network is specified, ranging from neurons to whole-brain regions or functions. How do you choose scale and how that affect choice of math methods. Depends on what questions you’re asking as to what scale you should choose. Time constants of homeostatic sleep for humans can be tuned to work. For rats homeostatic sleep build up too quickly so can’t compensate with circadian clock. Shows importance of time constants and that it can really give different effects and conclusions depending on what you think or use here. Could be order-of-magnitude difference between rats and humans. Populations of neurons can shift clock and likely mediated by electrical activity and firing in SCN. Can use Hodgkin-Huxley type models here. Circadian rhythm plays a role in sleep timing but not the ultimate clock! Feedback with other sleep clocks!
'''Marishka Brown:''' It is a mandated program but funding is not mandated. Largely instigated with push from circadian researchers. Trans-NIH sleep coordinating committee to build partners across all the different parts of NIH. Starting to build with NIAID. Studies on fatigue and sleep and performance joined together for studies. Connecting also with Human Health and Services. Circadian, disease, development, understanding are big parts of funded NIH sleep proposals. Turning discovery into health is motto. We scientists need to communicate better with those who fund us and general public to explain why basic science is so important to even be able to have something to translate into clinical applications. Need to do a better job to verbalize priorities to get funding behind us. Sleep disturbance is one of the best predictors of suicide. Lots of effects on heart and other functions. Correlation between sleep deficit and every type of risky behavior of adolescents. Public is convinced about sleep, just not academic medicine. Sleep and circadian rhythms are fundamental to health and life. How does circadian timing (NIAID) and sleep restriction affect immune response and vaccinations? Funding opportunity now. Ideas about chronotherapy. Does timing of organ donor for transplant need to match circadian timing of recipient or does it not matter? How does sleep deprivation affect it? What about sleep and health disparities compared with woman, URM, low SES, rural communities, etc. Why do these differences exist? (Not just cardio event in AA adults.) DOD strongly aware of importance of sleep and funding for research for it. Big upsurge in funding over last 3 years. Funding connections with cancer are expected to grow stronger. Silos of sleep research based on citation networks. NIH wants to break that down and create more communication.
'''Victoria Booth:''' Shortest time scales are bouts of minutes. (In principle, could this go down to even milliseconds because memories can form in that small of a time.) Wake durations are power law and sleep durations are exponential. Wake durations seem to follow Zipf’s law. Sleep duration follows random Poisson process like radioactive decay. More fragmentation in sleep bouts happens towards the end of the night. How strict should we be with definition of REM and SWS by voltage or time-scale and frequency characteristics across ages versus qualitatively similar behaviors. Randler et al. 2019 in Sleep Medicine for time in bed across ages. 3 to 4 months for baby to get used to day-night cycle. Sleep is more fragmented with aging. Trying to use fine time-scale modeling to predict changes in sleep across development and aging. Both wake and sleep bouts start exponential and random but become power law for wake around P15. Is this same as pre-term to term? Lesioning SCN you don’t see as much of the power-law behavior being established. By synaptogenesis P14 in rats is similar to about 3.5 years of age in humans. Younger subjects have longer tails than older subjects. Difference has a lot to do with NREM to REM transitions. Probability of waking up is much higher in older people so they sample the long-tail of the distribution and have longer individual waking bouts. LC develops early in fetus (Nakamura) and transitions/switches to alpha-2 autoinhibitory mechanism around P10 to P15. LC might be off in REM sleep. True in fetal rats but unclear in humans or other species? Diniz Behn and Booth (J Neurophysiol 2010) have physiologically-grounded model to investigate and explore this with and look at mean firing rate using ODEs. Stochasticity lengthens tail. Huge switch at 3 to 4 weeks to entering sleep in NREM phase. Can mode predict this?
'''Jerry Siegel:''' Sleep changes across species. Argues it’s driven by ecological factors. Argues sleep is adaptive in evolutionary sense. Perhaps animals are dying from repeated wakenings instead of loss of sleep for things like disk-over-water technique or experimental deprivations in general. Carnivores show little change in total sleep time with size. Recorded conditions aren’t much like natural conditions so how much does that affect sleep results you get. For example, temperature regulation. Example of platypus believed to have no REM and then they built platypusariums and there was lots of REM. Sleep changes a lot during migration. Amount of sleep in wild typically less than in lab. Elephant sleeps half as much in wild as in lab so about 2 hrs. Do they sleep at best time to thermoregulate. Bats sleep a lot (~20 hrs) but also have a very high metabolic rate. Go into hibernation though SWS that gets deeper and deeper. Could REM sleep be happening while elephants are seemingly awake and walking around? Might it be happening in only certain parts of the brain? Could you see how Mom responds to meat in terms of speed and orientation to determine how lack of sleep affects her. EEG basis for REM sleep but not based on brainstem recording. Call to include more temperature effects in models. Temperature and light are correlated. How to disentangle which drives sleep? Janet Best has paper on how temperature drive sleep cycles.
'''Geoffrey West:''' Body cannot have nearly perfect repair because of increase in entropy and second law of thermodynamics. But brain can be treated as not a closed system. It is an open system that draws from and drains off of resources of rest of body to keep nearly perfect repair. Lifespan and aging both about damage. Should be able to have formula in biology textbook for why we sleep 8 hours a day and live roughly 100 years.
'''Alex Herman:''' Subcortical REM more important in relation to number of synapses.
'''Bob Stickgold:''' More creative after sleep or dreaming. Because new creations are explored then or puts brain in better space or configuration to then do exploring once you wake up. How does this relate to patients with narcolepsy because seems like they have higher levels of creativity or lucid dreaming. Function of dreaming is not to solve problems as much as to explore different solution spaces and find what’s promising. Reminds me of trying to devise better computational algorithms for exploring large dimensional data or parameter space for theory. Maybe our brain has found good way to do this. Question is why does this require sleep to occur. Is it because it isn’t accessible to our conscious brain because done by some other part of the brain and takes lots of energy so have to save energy from elsewhere? If it’s so useful like this, why do we forget dreams? Maybe because their actual content isn’t useful? Or maybe we unconscious access to them but not direct access. Or maybe if you remember too well, you start to confuse dreams with reality and that becomes big problem.
I
'''To discuss:''' Could we make a grocery list of all irreversible processes mentioned in the workshop? Few that come to mind immediately: (1) Priyanga's idea on hot to cold invasion (2) Ecological succession. (3) Niche filling (4) something funky happens with the lottery model when there is a big mutation event (is there a name for this? Surely it is a generic thing that can happen in many other systems) (4) formation of interdependences / mutualism / specialization (I will mention in my talk tomorrow) (5) gene duplication (I couldn't follow all there steps here). Anything else I'm missing?
More specific thoughts on individual talks:
'''Otto Cordero:'''
Succession of species on hydrogel microspheres. Otto uses spheres with four kinds of nutrition. '''(1)''' why are species either specialist (able to digest only one type of sphere) or generalist (able to digest all types). '''(2)''' Why don't the cheaters (those who do not produce digestive enzymes) take over. '''(3)''' Prima facie, I would expect cheaters to have much lower detachment rate. They should just stick onto spheres and wait for the digestive bacteria to arrive. For the bacteria doing the work a better strategy is to detach quicker, at least before cheaters arrive. Is this observed in experiments? '''(4)''' is the interaction between bacteria indirect (i.e. they compete for the same resource) or do they secrete antibiotics or consume one other? '''(5)''' what is the role of diffusion lengths? The commensalist bacteria (those who do not secrete enzymes, do not compete for the main resource, but utilize the metabolic byproducts of others) gather whatever they can within diffusion length. So we can calculate the limit the number of layers of bacteria on a surface. For resources (e.g. dead crab shells) that are smaller than the diffusion length, the shape and size of the resource will also make a difference. Also calculable.
'''Pamela Martinez:'''
Markov process to describe the spread of pathogen with multiple serotypes (a kind of SIR model). How to differentiate between different models with sparse data. Data could be equally consistent with randomly connected states or even a single Poisson process with appropriate mean. A good suggestion during the talk: generate synthetic sparse data using the model, pretend the data is real, and estimate model parameters. Do they have a similar value? '''(1)''' why does the efficacy of vaccines not show up in the population data (they do make a significant difference in controlled studies). '''(2)''' why does the vaccine work on some countries but not the others '''(3)''' Rotavirus somehow interacts with the gut microbiome. There is some literature that shows that vaccines work for people with microbiomes of the "european kind".
'''Annette Ostling'''
Starts with Lotka-Volterra type fitness function. Species are assigned traits between 0-1 and and the interaction matrix is structured such that species with similar traits antagonize each other. This is done with a gaussian kernel in the sum. '''Questions:''' '''(1)''' what determines the number of clusters. Can I use Turing analysis to solve this analytically? '''(2)''' Can I view phylogenetic branches as "clusters"? e.g. animal kingdom, plant kingdom etc. are, in some sense, clusters. And then, there are sub-clusters within these clusters, and sub-sub clusters. What feature should be added to the model to obtain sub-clusters. '''(3)''' Given an empirical distribution of features (within a species or within multiple species supposedly filling a niche) how do I distinguish between environmental filtering vs exclusion?
'''Priyanga Amarasekare:''' Her argument is, species respond to temperature in an "asymmetric way" (specifically, you hit a wall at high temperatures, but the negative response to cold is more gradual). This leads to an irreversible flow of species (via mutant invasions) from hot regions to cold ones. '''Comments: (1)''' I like the idea a lot, very plausible. Here is my alternative (and quite possibly false) point of view: A high population is more evolvable, because there will be more mutants/innovation. Warmer climates have higher biomass (just because it receives more energy) and will therefore generate viable invaders at a higher rate. Maybe.
'''(2)''' She had some discussion about constraints vs selection. It's a possible to dichotomize, but I view these two things as one thing. A constrained region in phenotype space is just one with fitness=minusinfinity, so no one visits there. Possibly just a matter of semantics, but in any case, I don't see how a constraint implies irreversibility.
'''Jacopo Grilli:''' Three-body interactions surprisingly stabilize the community (unlike those with two-body interactions). I found this surprising because the model with three-body interactions is really an effective model of a two-body interactions. e.g. species A,B,C come together; first A interacts with B, then the winner interacts with C. (and you symmetrize this, because sometimes first A interacts with C first). As such, the outcomes of this model should reduce to a Lotka-Volterra model, (with specific structure, under specific conditions). However, I was not able to figure out what this structure is, and what the conditions are. Whatever the structure and conditions, the stability of the system with three-body interactions should not be a surprise if the equivalent Lotka-Volterra equations are also stable. Either way, I would like to understand this better.
'''Greg Dwyer:''' The viruses that infect the pests have multiple DNA's, so I thought that might give rise to an interesting cooperation/cheating dilemma, similar to the one we see in sperm trains. Also there is an interesting three-species coevolution going on between the pests, and the virus and fungus that infect the pests.
Greg likes things he can measure and doesn't like discussing the meaning of life. But then he was converted, and found the meaning of life. Turns out meaning of life is measurable after all.
H
''Overall thoughts: A system can take a certain # of hits. A lot of failures are linear with age on a log scale. Why? Resiliency? Due to redundancy? What explains variance in a population? Natural selection has acted to limit physiological decline to the point that maximizes reproductive success (balanced against costs). So risk of physiological failure are not zero during reproductive years, but are minimized, certainly relative to competing risks for most of our (or other animal's) evolutionary history.''
<u>Bernie Crespi</u> presented on how selection works stronger on the weaker components. Slow, invisible weaknesses increase over time. Can engender exponentiality of decay. Or is it that damage and mutations accumulate over time? Risk of failure is related to the degree of complexity in a system, and tightness of coupling between components. For example, power grids are tightly coupled but not hugely complex. Aircraft are both coupled and complex. Mental disorders represent alternative attractors in a highly coupled system. For immune system, autoimmunity, chronic inflammation, cytokine storms represent alternative attractors potentiated by tight coupling. Both can have diseases associated with over-defense. Both do not trade off well with other systems. Immune system is under greater positive selection, relative to brain system which is more stabilizing. ''Thoughts:'' ''I would ask how the failure rate of systems, complex or not, is impacted by environmental disconnects. For example, failures in the immune system (like autoimmunity) may be common today, but were not so common in ancestral times (re. hygiene hypothesis).''
<u>Dario Valenzano</u> described how relaxed selection shapes the rate of aging across species. Killifish get tons of cancer late in life; almost all die with cancer. Not clear if they die of cancer, just with cancer. Some species live in areas where water is available for a few months of the year. Have drought resistant embryos. Can desiccate embryos. Hatch when rains. 5-6 rounds of spawning. Wetter areas – longer lives. Did a lot of genetic mapping. QTL of longevity. Can do crosses and single alleles can have big effects. Show heart, skeletal, and many other aging phenotypes. See evolution of life spans multiple times across killifish species. So can filter out phylogenetic relatedness. Short lived (annual) species have a larger genome mostly due to expansion of transposable elements and introns. A lot of LINE expansion. Probably due to reduced purifying selection. Much more of coding sequence is under purifying selection in non-annuals. Populations evolving in dry environments have smaller Ne - even within the same species found in dry and wet environments. See expansion of mitochondrial genomes in annuals too, with a higher mutation rate (again consistent with weakened selection). In fact, the mitochondrial polymerase is more error prone. ''The question that I would raise is WHY is selection weakened? Clearly, a smaller Ne would do this, but why is there a smaller Ne? Is there a bottleneck with each season where only a small fraction of dried embryos survive and reproduce again?''
<u>Shripad Tuljapurkar</u> – As we get older, get more response to similar challenges. ''My thoughts - but do we? Or do we simply return to homeostasis more poorly? In fact, sometimes responses are exaggerated and prolonged (such as with inflammation) in old animals.'' Escape from “well” for a phenotype is dependent on depth of well, size of fluctuations, and multidimensional space. Old age involves shallowing of healthy well and deepening of sick well. ''Thoughts: the well may also be broader, with a greater diversity of phenotypes (cellular) for the same cell type. Relaxed purifying selection, as well as epigenetic drift, could contribute to this.''
<u>Sabrina Spencer</u> discussed how replication stress and DNA damages induces p21 in mother cell (somatic cells in culture), leading to increased p21-high offspring which stay quiescent longer. And these cells are stress resistant (survive better D dam). See much less pausing at early passages of foreskin fibroblasts relative to late. ''Thoughts: It will be interesting to see how these dynamics contribute to tissue aging. Also, in a competitive tissue do these same cells with DNA damage get purged? There is evidence from flies and mice that that damaged cells can be pushed out by healthy neighbors. From discussions during hike, it would be great to have a method to identify cells with critically short telomeres in a population (such as by marking gH2AX colocalized with a telomere binding protein - but not feasible for living cell...).''
'''<u>Summary of Day 1:</u>'''
Themes: 1) Basins of attraction, 2) Failure (gradual, rapid, can we predict?), 3) Underperform or misperform, 4) Tradeoffs and pleiotropies, 5) Networks, not discrete compartments. 6) complexity vs simplicity; complexity vs complicatedness, 7) Don’t make a part that consistently outlives the whole.
'''<u>Day 2:</u>'''
<u>Morgan Levine</u> – DNA methylation landscapes in aging and disease. Changes on the molecular level propagate up through networks, increasing risk of disease. Every level above can tolerate some level of failure below. Biological aging proceeds, phenotypic aging, which proceeds functional aging. Biological age can be different from chronological. Use epigenetic clocks – methylated CpGs. Measured at 353 CpGs throughout genome. R=0.98 for chronological age. Do correction for <12 year olds to account for rapid change in early life. Developed “phenotypic age” predictor of aging related mortality based on clinical measures, which can predict disease. Had almost 10,000 individuals aged 20 and up, with up to 23 years of mortality follow up. Combine 9 clinical markers (albumin, CRP, MCV, etc). Incredibly predictive; 1 year increase in biological age coincides with a ~10% increase in risk of CVD, cancer, diabetes and lung disease. Comparing top 10% to bottom 10% of a group shows ~2x difference in HR. Most people will fall within 5 years of their chronological age. And difference remains stable. For all cause mortality, get about a 1.05 HR (per year of change in biological age, I think). Correlations with DNA methylation clock too. Horvath clock corrects for differences between tissues (there are differences). Developed transcriptomic signatures, correlated with DNAm clock. Cellular respiration, lipid oxidation, and nucleotide metabolism correlate well with biological clock. ''Thoughts: how do changes in DNA methylation correlate or not with changes in DNA mutation accumulation? Some data from Sanger indicate that the increases in mutations from birth to death are linear, but epigenetic changes show a sharp increase during growth/development periods. Why?'' ''How can we take these results and make them useful for the individual? It's not enough to just tell someone they're going to die earlier than their chronological age suggests, but what could be done about it?''
<u>David Schneider</u> – measuring resilience of hosts to disease. Can measure tolerance to an infection. Can we predict trajectories for individuals? Not yet, but can for populations. Looking at tolerance across 150 mouse strains from crossing 8 different inbred strains. If plot NK cells by RBC, get loop. Can figure out where someone is on a “disease clock”. Can see wave of parameters, as one rises while the previous falls. IFNg is first, gamma delta T-cells last. Kynurenine comes on when animals are sick. Orotate is late. Can look at curves for different mice strains (some die, some survive); 29 metabolites correlate with resilience (can predict who will die and who won’t). If arginine bottoms out, the mouse will die. And arginine really is critical for surviving infection. Kynurenine pathways is also critical (as it is for aging). Parasite load correlates poorly. If treat with anti-parasitics before d6, reverse on curve. But not if after. Also showed that glucose can prevent deaths from infections (provides energy). Our mouse houses are at 20 C, while prefer around 30 C. Get hypothermic upon infection instead of fever. Check out his blog: Phasecurveblog.wordpress.com
''Thoughts: how do we alter responses so that recovery is favored over failure (death)? Can we manipulate metabolites? His and perhaps other labs seem to be doing this.''
<u>Marten Scheffer</u> – ecology of resilience. How quickly is the return to normal after perturbation? He studies lakes. Sometimes they don’t recover. Due to severity of insult or the brittleness of system. Catastrophe theory on tipping points. How to know a stability landscape? For example, a tropical rainforest can be become less stable if there is less rain. If plot tree cover with rain across Earth, not a simple correlation, but discrete clumps – attractor states (dips in landscapes). See for bacteria in human gut with age. Basins of attraction are shallower in old age – smaller disturbances can push out to new state. Greater resilience can reduce cross talk between systems (negative cross effects). Can find early predictors of climate change, depression (mood), etc. Do we need to make perturbations in humans to predict future health? ''Thoughts: how do manipulate resilience? Can we increase it or at least prolong it later in life?''
<u>Wrap up</u>: Develop frameworks for understanding biological failure. Can we derive concepts that inform models? Bernie – can connections be made between larger theories (Antagonistic Pleiotropy and Mutation Accumulation) with molecular changes like in epigenetics. What are patterns of change? Why do they change? Feedbacks maintain homeostasis and restore it post injury.
Hmmmm. Why is so much of what bad happens to us as we age exponential?
Our charge: Frame what we agree on, what are the big questions, and what we could do to get answers (with unlimited resources) and what can we not answer?
1) Try to come up with fundamental and important outstanding questions in term so biological failure?
2) What are the kinds of connections that we can make between conceptual approaches that will allow us to work on these?
(1) General problem/question
DNA/RNA/proteins - cells - cell-to-cell interactions - tissues - organs - organ systems - whole organism - groups of related organisms as in colony - community including non-relatives
how/why does senescence occur at these different levels, in the contexts of antagonistic pleiotropy and mutation accumulation? does balance of early benefits vs later costs vary across these levels and systems?
does the complexity/stability/entropy of the system matter? how?
why does senescence accelerate exponentially in humans - synergisms between failure types/levels? why up so fast around age 60-65 in particular (later for women)-becoming less useful as a helper/contributor/transfer-er then, and are being 'replaced' these roles by next generation?
(2) Caloric restriction - reduces deleterious aging effects on health but does not lengthen lifespan if started after adulthood - why not? increases investment in maintenance, reduces growth and reproduction, based on energy, thus mitochondria
all cells use energy - does energy restriction impact some cells/systems more (eg brain if not fully protected); or are all systems just ramped down - how does this impact on disease risks, exactly - do diseases require more energy? or are fewer mistakes/mutations made?
(3) Single cell talk p21. Is this work key to cancer biology in that tumor suppressor functions being lost leads to all cells just keeping reproducing? can one show that senescent cells are less likely to transform, using this system?
(4) Idea that senescence is an emergent property (of defense against worse failures) appears quite important to me
(5) Comparing biological failure between organisms and higher 'levels' (societies, civilizations, ecosystems, etc): how might societies/civilizations age in same way lower-level systems do? maintain current system even when its benefits decline, due to environmental or other change? external causes of increased weakness of units at different levels? Fit with Jared Diamond's ideas about Collapse of civilizations? (what were those need to check).
Parallel causes, biology and higher levels - complexity and coupling of components? Too much investment in defense? (Russia 1991)
DAY 2 questions
how do CR genes relate to antagonistic pleiotropy?
what are roles of germline mutation in relation to somatic mutations in senescence?
intelligence/brain size related to aging/senescence both within and between species (and related to CpG calendar/clock methylation) - why? do immune systems have 'intelligence'? (measure of their efficiency of functioning in some manner).
why length of post-reproduction life vary (esp females) - fewer stresses when young? more resilient senescence system? other? related to intelligence (more than pre-menopause lifespan)?
age around 60 represent tipping point into accelration of aging effects/disease risks? synergistic failure across systems?
adaptive landscape model - flatten with age as selection weakens, lower adaptive peaks? mutation moves you downhill easier?
prioritization of organ systems with age - how vary - costs of degradation is each system? brain, immune at top.coindcides with ability to trade off across systems, higher->less prioritization
aging and loss of information in networks, higher entropy, protein folding, feedbacks in networks, how measure these changes, chaperones, perturb system like insulin challenge test in pregnancy
DAY 3 thoughts, questions
How well does theory interface with data in study of aging?
Agendas of proximate-mechanisms, biomarker geroscience researchers how much integrate evolutionary or other theory insights, how useful will they be for advancing theory?
Empirical ideas
Crispr on CR genes then analyze behavior of cells/tissues across age
CR genes for maintenance (rheostat) how much trade off with growth and reproduction genes, strong theory predictions here
[aside: do empiricists who focus on specific function tend to not think about tradeoffs (in evolutionary contexts) because they involve competing/beneficial but incompatible functions?]
DIfferent organ systems ought to senesce at different rates
W
* Add information/discussion about non-brain "functions" of sleep (e.g., metabolism, immune function) with their time scales
* Add Other brain "functions" of sleep to include are: mood/psychiatry; objective neurobehavioral performance; "cleaning" of toxic substances (glymphatics); neurochemistry
* What are target/pivot points for changing some aspect of sleep?
* How to define and then study sleep "need"? Is it different for each function? with age? How does it interact with sleep duration or timing of sleep?
* Feedback between sleep and circadian rhythms; not just one way circadian-> sleep
* Two comments from discussion:
** Comment: Frame the “needs” of sleep not as competitive with each other but as parallel.
** Comment: Not all functions of sleep bring you back to where you were before; some are to bring to a new state (e.g., learning, immune)
+
H
* Considerations of time-scales. Where/how is time encoded?
* With time, new structures emerge, network connectivity changes
* Can hyperactivity that optimizes growth and development contribute to aging (antagonistic pleiotropy)
* Peto’s paradox—what things besides the number of cells, which should influence probably of neoplastic transformation, confer differences in species-level cancer risk?
* Only the intercept, not the slope for the age-specific rates of lung cancer differ between smokers and never smokers.
* May not be intrinsic damage accumulation and instead may be the difference of the “aged” vs. young environment which differentially selects for oncogenesis
* Clonal hematopoiesis could be facilitated/accelerated by the old versus young environment?
* How does this fit in to the findings from heterochronic parabiosis? Would changing the environment of the niche from old to you or vice versa alter 1) clonal hematopoiesis, or 2) probably of neoplastic transformation? What is the impact of HRAS in young versus old environment?
* Aging may be both the small perturbations to the system and the reaction to perturbations in neighboring systems for which one depends on. With a healthy neighbor idea, a small perturbation may not matter if the interacting system is functioning well, but ass the systems all go down hill together, there will be feedback that will cause an acceleration in the rate of multi-system declines. This may account for why things that change linearly with age, produce exponential morbidity/mortality cures at the level of the population.
* CR is changes a lot of pathways that together produce both lifespan and health span extension.
* One can start to map the CR response to systems-level (or network) changes involved in a variety of pathways.
* Interesting overlap with the genetic architecture of insulin resistance
* The same pathways that come up in CR are also the ones we are finding relate to the epigenetic clocks in humans. Also evidence that CR decreases mouse epigenetic age.
* Wells that define healthy vs. sick and transition out of a well depends on depth, curvature, and the size of the fluctuations.
* A hidden Markov is a good way to model this because the transition to a given well/state depends upon the current state—doesn’t matter how you got there just the parameters above (I would also add that there are likely many well and proximity to a well influences probability of transition).
* I would ask whether the overall landscape is changing, or whether the available landscape (what is in your vantage point given where you are) is the only thing that changes. If we think of the wells as discrete states, they shouldn’t actually change, just your probability of sampling a given space given your current space will change.
* Can you also model entropy using these wells? Basically you will have a tendency to move lower in a landscape towards equilibrium and moving uphill requires energy input.
* Does it make sense to consider single-cell data and investigate the probability distributions of having a multi-dimensional profile as a function of age (could use something like Mahalanobis distance).
A
- General disagreement on the definitions of aging and senescence. It does not seem a purely semantic problem, but reveals the lack of universally accepted essential conditions to define the aging process.
Example: Role of asymmetry (necessary condition or not) in the cellular division process for the emergence of aging.
- Discussion on the more appropriate measure of time to observe aging of living things.... The intrinsic timescale of different organisms (for example doubling time in fast-growing conditions) and the timescale fixed by the environment (e.g. feast and famine cycles) should be taken into account.
[interesting parallel with the approximately constant lifespan of mammals if measured in heart beats shown by G.West]
- In several different systems the hazard rate (for cell division or for death) shows an exponential increase followed by a plateu (e.g. in mortality rate in E.coli shown by Uli and from Lindner group) . Different theoretical models can reproduce such a trend. Is there a way to do model selection?
- Need for measurements of metabolic rate in bacteria in order to build a complete phenomenological theory of resource partitioning and its relations with growth and mass. An extension of existing theories built on data based on balanced exponential growth probably need to include survival and the cost of maintenance... +
C
- dynamics: if we think at aging as the approach of critical transition we have specific predictions: critical slowing down, flickering, etc
- information: where is information in this dynamical / tipping point picture? Do we need it?
- scales and information: at what scale should we look at the brain to "explain" the interesting (information) phenomena? +
-applying different approaches in complexity towards understanding age-related changes in brain organization (e.g., thinking about flickering, network 'clogging')
-prospects of expanding/linking our work to additional scales of analysis +
H
<u>'''Day 1.'''</u>
<u>Bernie Crespi.</u>
Accidents (failure) are inevitable (normal). Challenger disaster, Max 8 jet. Risk of failure depends on complexity of the system. Tight coupling vs modular. Tight coupling – butterfly flaps wings causes a storm elsewhere.
Plot of tightness of coupling vs complexity. Brain and immune system go in top right corner. Is biological failure mediated predominantly by the brain and immune system?
Can make this plot at top right corner at the cellular level too.
<u>SLS:</u> how to test resilience of cells? Can we come up with a set of perturbations and appropriate responses? A stress test for humans before a surgery, or for populations of single cells?
Disease and senescence anti correlate with intelligence. Correlation between intelligence and lifespan is mostly genetic. A result of having more ‘good’ genes. Peter Visscher lab 2016.
Mental disorders represent alternative attractors due to tight coupling. Schizo, depression, autism.
Neurons have to be well-defended bc they are long lived and aren’t replaced.
Neuronal stress: age, apoe4, infection, energetics, insulin resistance, sleep deprivation.
Bodily senescence is mediated by over-defense and inflammation.
Failure due to bad tradeoffs.
Is senescence due mainly to defense against death? The defense ends up killing you.
The immune system is what is keeping people alive in the face of infectious disease.
Loss of coordination among organ systems causes health risk. They are networks. brain is made up of many different kinds of cells including immune cells.
<u>Dario Valenzano.</u>
Relaxed selection shapes the rate of aging across species. How can we reverse time and intervene in aging?
Killifish live 4 months.
Some strains live 15 weeks, others 30 weeks. Cross them and get genetic maps controlling lifespan.
<u>James Degregori.</u>
Why do we get more cancer late in life? Not the accumulation of oncogenic mutations. Loss of tumor suppressors actually reduces renewal of stem cells. but past authors focused on increase in cancer.
Investment in tissue maintenance during youth.
New model of multi stage carcinogenesis where a mutation has a negative impact early in life and a benefit later in life. Bc a stem cell is not well adapted to an old lung and thus will evolve. The same mutation can be maladaptive early in life but adaptive late in life.
Bad cells are poorly adapted to their environment and are pushed out.
It’s not the cell, it’s the mismatch between cell and environment. Adaptation is more likely bc the system is not working.
Senescent cells are a huge part of the environment.
Adding another dimension to somatic evolution.
IL37 transgene is anti-inflammatory. *
Inducing a genetic translocation with crispr in young mice did not cause tumors. But it did if you induce in old mice. But not if you suppress the immune system.
Quality control goes down late in life.
Lawrence Loeb showed that increasing mutation frequency by mutating DNA Pol delta’s proofreading function does not result in increased cancer.
We are loaded with oncogenic mutations. We are not avoiding cancer by avoiding mutations.
Can capture the effects of caloric restriction by increasing autophagy.
<u>Rozalyn Anderson.</u>
Caloric restriction (in the absence of malnutrition).
RNA processing is different in the caloric restriction group.
Push pull of metabolism vs growth… gets disconnected with age.
Increase expression of PGC1alpha in mitochondria by 1.5x to mimic caloric restriction.
Wound healing is slower in CR animals. They would not be vigorous in the wild. They do things more slowly. The animals are smaller so there’s a growth effect of CR.
Constant vs periodic caloric restriction? Intermittent fasting. Fasting and resilience research is growing.
<u>SLS</u>: How well can you capture the benefits of continuous caloric restriction by doing 12hr fasting?
The periodic '''ketogenic''' period is important. Can now look for panels of molecules and patterns of change.
<u>Shripad Tuljapurkar.</u>
Back and forth between evolutionary timescales and one person’s lifetime.
Assume we are the same today as the Romans were.
When we are young we can handle challenges (like missing the bus). We have a muted response to challenges when we are young. Older people worry about things and have higher amplitude responses.
People get less homeostatic as they age.
Recovery is harder as you get older.
Menopause has to do with the number of reproductive follicles that you are born with. (?)
Grandmothers; learning and wisdom. Theory of transfers. Can transfer care, knowledge from old to young.
How much added longevity can we explain from transfers? 15 years. 50yrs to 65yrs old.
<u>Questions I got from the audience after my talk:</u>
· Is high-p21 CDK2-low state a trap? Bc p21 goes up and up, suppresses CDK2 more and more, which makes it harder and harder for a cell to escape and re-enter the cell cycle.
· Would every cell go through the CDK2low state at some point? Can you develop a sensor that would turn on once a cell goes into that state once. (sc)RNA-seq to identify features of CDK2low cells that would be long-lived or permanently on.
· Does % CDK2low cells increase as cells become senescent? Use primary cells at increasing passages.
· Have you normalized p21 and 53bp1 curves by CDK2inc? Set CDK2inc trace to 0 and look at points of convergence with the other 3 subpopulations.
· Similarity to yeast aging and asymmetry of mothers-daughter division?
<u>Barbara Natterson Horowitz.</u>
High adrenergic cardiovascular events = stressful events.
Vasovagal syncope = Fainting. An external event (like getting blood drawn) triggers slowing of heart, vasaodilation, underperfusion of the brain.
Alarm bradycardia. Loss of tone and fainting is a life-saving event when animals are being hunted. Playing dead. It is primarily in juveniles.
In adults, stressful events lead to fight or flight, not fainting.
A developmental response. Until you’re able to run away successfully, there’s a parasympathetic response. 13 years old is age of highest fainting
High adrenergic events cause heart attacks. Spike during Northridge earthquake.
Heart rate variability.
Restore dynamic properties of autonomic nervous system with exercise or mindfulness meditation.
Trapped shorebirds – 10% have heart attacks (and die?) after being trapped with a net. Some species fare worse than others – can you use the differences to predict which patients will have heart failure at the next stress.
<u>General discussion.</u>
Themes that have come up:
Variance.
Potential wells. Hysteresis. A to B is not B to A. light stress vs strong stress.
Resilience / homeostasis.
Can you see how frail a person is before a surgery? Don’t protect the elderly from all stress, give them some slight challenges and some exercise.
Measures of potential wells. Cell-to-cell variability. Maintenance of a proper phenotype.
How can we compute/quantify potential wells?
Define health: Physiological, psychological, social.
What is surviving? The ecosystem.
Scales: Genes, molecules, cells, tissues, individuals, populations, ecosystems, civilizations
Why do cars age and collapse?
Tipping points.
What is aging? Relationship to immortality?
<u>'''Day 2.'''</u>
<u>20190409 Day 2.</u>
Themes that came up again and again:
1. Basins of attraction. Wide, narrow, deep, shallow. For sick and healthy.
2. Failure. Gradual, rapid, can we predict?
3. What is failure? Underperform, misperform
4. Tradeoffs. Investment in critical functions could be associated with aging.
5. Everything is connected as a network (eg organs)
6. Complexity (many knots) (vs complicated – many folds)
<u>Let’s define aging:</u>
Programmed development. As well as non-programmed stochastic decay.
Process of aging vs pathology of aging. Separate disease-related aging.
Chronological age vs biological age
How does body protect itself and make itself new again?
It’s an open system so you just age.
Lack of rejuvenation and regeneration.
But a 40 year old woman can make eggs that produce a 0 year old baby.
Aging is entropy
Aging vs longevity
A series of repair mechanisms for wear and tear.
Cataracts and fibrosis are pathologies of aging. These are repair mechanisms that allow longevity.
Aging is suboptimal adaptation to environmental stressors
Aging is an active response, not passive.
Don’t forget fitness.
<u>Morgan Levine.</u>
Geroscience. Failure starts at molecular level, each level up can tolerate some issues.
People age at different rates. Can we quantify biological age.
Epigenetic clocks. CpG methylation. Usually turns down transcription if it’s in the promoter.
Horvath clock. 53 different tissues. Many post-mortem. Measure 353 CpGs across the genome and get a strong age predictor.
Hannum clock was trained on whole blood.
Uses machine learning to minimize the residual of the fit to a line. Want some residuals since an r=1 gives no information (you wouldn’t be predicting their age, it would just be their age).
It’s exponentially changing under 15 yrs old, then linear.
Get phenotypic age, then try to infer biological age.
10k people. Take out accidental death and HIV.
Q: Why focus on all-cause mortality? And not likelihood of a specific disease? Bc age increases your risk of all diseases.
We have multiple biological ages across different tissues.
Most people fall within 5 years of their biological age.
People are stable. Age at one time point is predictive of age 9 years later. People age 1 year per year on average. People don’t reverse.
Levine clock. Can predict age at menopause.
Most of the CpGs do not overlap across the 3 clocks.
Want to cluster the CpGs into module to figure out what they represent.
<u>SLS</u>: do you do any longitudinal studies? Do take your own tissues over time and predict age?
<u>SLS</u>: have you compared these clocks to telomere lengths?
<u>David Schneider</u>
Measuring the resilience of hosts to infections by mapping disease space.
How sick will you get with a given load of pathogen?
Tolerance to microbes is a population-level thing. Can’t do with one individual.
Health is a stable state, death is a stable state, but sickness is a set of many transient states.
Resilience: 1. How stretchy the system is. 2. How far you can go before you break.
With age, the phase diagram might change such that a state that was survivable when you are young is no longer survivable when you are old.
Measure cytokines, parasite density, NK cells, RBCs, etc over the course of the infection.
Could potentially predict where in the trajectory a kid is who comes to the clinic with malaria.
IFNG is first, gamma-delta-Tcells are last.
29 metabolites before infection predict which mice will survive.
Looking for critical transitions.
Q: Do animals have to always go all the way around the curve? No, if you treat them with drugs before day 6, the curve reverses, but not after.
Can rescue ketogenic mice from death by providing glucose.
Now want to do it all in humans.
Q: have you done this in old vs young mice? Yes, in 70 week-old mice
<u>Marten Scheffer.</u>
Resilience. How quickly things go back to normal.
In ecosystems, sometimes they don’t recover from a perturbation. They become brittle.
Landscapes are a useful metaphor.
Tipping points – 2 alternative basins of attraction.
How do you know where you are in a landscape?
Look over every square kilometer at tree cover vs amt of rain. Scheffer Science 2011.
Forest is stable, savannah is stable.
Abundance of a particular gut flora increases with age.
Systemic resilience of humans and other animals. Scheffer PNAS 2018.
Can we infer fragility? Climate collapse, societal collapse?
Generic early warning signals = indicators of resilience
Critical slowing down happens at mathematical bifurcations. If you push a ball and the slope is shallow, it takes a long time to return to normal. Fluctuations are longer, are more correlated over time.
Resilience in mood if a bird shits on your head. One hour later, are you happy again or still down?
All systems, human, earth, are continually being perturbed.
<u>SLS</u>: Can we make use of low-level fluctuations in single cells and compute autocorrelation to calculate resilience? Can we use this to measure cellular age? **
Measure resilience by measuring small fluctuations over time. Systems close to tipping point show:
1. Larger fluctuations
2. Stronger autocorrelation
What the other working group talked about: Measure balance on a balance plate (elderly do worse); Measure blood pressure during rapid standing from sitting (blood pressure slower to return to normal in elderly). Once you are frail, it’s already too late.
How can you predict from autocorrelation how far you are from tipping point? You cannot measure the distance to the tipping point. Bc it depends on chance.
