 Okay, well, thank you for the invitation. So what I'm gonna talk about is not just dynamics, but maybe try to bring a little ecological principle into the study of dynamics in the context of the human microbiome. And obviously most of my work has been centered on the human vagina. So I will bring examples that are derived from our study of the dynamics in the human vagina. But before I start, I just also wanna say that it's very important to understand that especially when it comes to those principle, and I'll reiterate this later, that not every body site is gonna be governed by the same principle. And I think we've seen this many times, comparing diversity, for example, the cross-different body site, and the vagina appears to be one of those sites that appears to behave differently, with low diversity being something that's good, when in a gut we just heard low diversity appears to be associated with something bad. So that said, before we go on and discussing dynamics and so on in the psychological framework, we might wanna start discussing a little definition. So in the context of ecology, dynamics is only one of the most important central theme to study microbial community ecology. So the dynamics of microbial community can be defined as a temporal and spatial changes in community structure as well as or function. So every time we talk about change, this idea of stability comes along. And I know some of it might be a little redundant with what David just talked about, but stability and instability has been defined extensively in ecology, but also was quite a diverse set of views to it. In 1997, those author catalogued over 167 definition of stability just in the field of ecology. But pretty much they could summarize this that stability encompasses three principle. One of them is this idea of resistance or what we call also constancy, where a community pretty much is staying essentially unchanged over time, as well as resilience, which David has talked about, where a community can be disturbed but returned to a reference state. And the last one is what we call persistence. And persistence, it's when the community pretty much persists in one ecological state throughout time. So one reason that there's so many definition of stability is certainly because they're not all asking, people using stability are not all asking the same question. But pretty much you can kind of summarize this by saying that stability, if you want quantifies the extent to which a community stay the same. And I think that just simplifies a lot for this work. So stay the same over a long period of time. And of course it depends on time scale. So if you're talking a lifespan, if you're talking a few years or decade, all this is gonna change. And it can stay the same over time or in the face of some disturbance. And I think David has touched on this quite extensively. So there are many, many ways by which you can measure different metrics that you have to measure stability. And I listed a lot of them. And again, might be redundant on what David just said, but depending on the metrics that you're gonna use to establish and define stability, you might get different answer. And you might have some community that appears to be stable given one metrics, but given another metrics are somewhat unstable. And this was just a paper by Jeff Gordon's lab recently that showed that by just introducing a probiotic, the community is not changing in composition, but its activity, its gene expression of the community is dramatically changing. So again, so it's very important in terms of the metrics that you can do. So either composition type metrics or maybe some more functional aspect of the community. And composition both include doing gene survey to get to relative abundance as well as some of the work, for example, by Dave Frederick in the vaginal microbiota that looks more at specific species, QPCR, where you really get to true abundance of some of the members of that community. So disturbance is another very important concept in ecology and it's linked to resilience. And David mentioned actually the second part of the definition of disturbance that last year where you can actually kill or displace one or more member or population of a community and that gives opportunity for other to take over the space. But there's other part of disturbance. You can disturb a community by, for example, removing a nutrients, changing a substrate of ability or even the physical environment of that community will have an impact on community structure. So there's many ways you can have disturbance. And resilience, of course, is the amount of disturbance that an ecosystem can actually withstand without changing itself, organizing processes and function. So ultimately people think that community that have a fundamental difference in species composition and structure and function certainly will differ in their level of resilience. So how can you kind of model this, all this together, this concept of disturbance and resilience? So if you think about disturbance, they can have certain level of intensity. They can have certain level of frequency and duration. And the combination of those two really define the disturbance. A community resilience can be, if you want, defined by this threshold line here. So a community can be disturbed, can be changing in structure or function and stay within this space that's defined by this arc here. On this side. But as soon as the intensity and the frequency, the combination of those two is high enough, the community can move into a disturbed state, kind of one of those other valley that David mentioned. So in the disturbance that we talked about, they can be actually imposed by human. And that would be, for example, our hygiene, diet, behavior, antibiotic use. And if you wanted to, it's our life history. And all those disturbance are gonna influence our community, microbial community. But they also include normal biological processes. For example, menstruation in women can have an effect. Hormonal variation over a woman's lifespan actually can have major effect on the vaginal microbiota composition and certainly function. As well as differences in changes in immunology, your immune status over time. So one thing we're trying to understand is what are the driver of change and resilience in a microbial community? So in order to understand dynamics, we have to study community over time. We have to be able to sample the community over time. There's been many studies that have done this and those studies have shown some very, very important results and finding. But they've all fell short pretty much, most of them have said, fell short because they either sampled very sparsely over a long period of time, or they sampled very few subjects very frequently. And the study that, for example, David mentioned earlier are some of those examples. While they're very important in driving hypotheses, it's very hard to generalize the finding to the larger population. And really, the reason those were done with those kind of study design is that they were really limited by cost and feasibility of sampling, for example. So here I just want also to say that some studies that studied, for example, the gut, the principle that are derived might not fit, again, what we might find and study, for example, in the human vagina. And the other thing is that it's very important, most of them actually looked at composition. And a lot of time the composition analysis is done at very low taxonomic resolution, often at the phylum, class, and order. And I think it's very important, and especially in the human vagina, to go down to even lower resolution species, and as you might see later in the talk, potentially even strain of certain species. So now let's move on and look at the dynamic of the human vagina in reproductive age woman. So we know that the vaginal microbiota is changing over time, through a woman's lifetime, quite dramatically, but for this particular talk, we're gonna be focused on just the reproductive age period. So recently, about two years ago, we published a paper that defined five different type of vaginal microbiota. And I represented them on this PCA plot. Five of those type, four of those type, are dominated by one species of lactobacillus. So we have lactobacillus chryspatus, lactobacillus sinners, lactobacillus gastri, lactobacillus gensenii. And the fifth one, which has now been divided into two group, actually define a community that does not have very large number of lactobacillus and has a lot of strict anaerobes. So this study kind of highlighted, first, the importance of lactobacillus as potentially a keystone species in the vaginal environment, but also that in asymptomatic women, we can also have community that don't contain lactobacillus. But this cross-sectional study did not give us any information on how a vaginal community can actually change over time, and even if it changed over time, or is it fixed in one time and it stay constant like we've seen, for example, in the gut studies that were presented earlier. So to address this question, we actually perform a study where we enrolled 160 women, and we did what we call a prospective longitudinal study where we collect sample every day, in this case, has women go through their 10-week period, and we collect everything, and hoping that suddenly some condition will happen and we'll be able to study whatever happened before, during, and after all those different events that might happen in the vaginal microbiota. So those samples were collected. We had a study that aimed at not only looking at DNA, but also RNA and also metabolomics and proteomics as well for future use. We had very extensive daily diary and a weekly visit at the clinic. We had clinical assessment at enrollment week five and week 10 of the study, and we did this study using power sequencing of the V1, V3 region. So we did the study for 50, we were able to generate data for the first 50 women. And when you start looking at this longitudinal profile, we can kind of cluster them if you want into three main categories. The first categories are women who appears to be stably colonized pretty much by one species of lactobacillus, and mencies, as indicated by those red little dot, they are the only event in those community that tend to induce a change in the composition of the community. So in this case, it's lactobacillus chryspatus is dominated at all, is dominant at all time. In this case, in orange is lactobacillus innus who's dominated at all time. Then we have a second group of women, which I don't know if you can call this stable, but somewhat over time are colonized by a wide array of strict anaerobes, but contain very, very little lactobacillus, and you can see a little lactobacillus innus here and none into this woman. But species that are actually been associated with, for example, bacterial vaginosis, this large green one here is garnalive vaginalis, this blue one is atopobium vaginine. So there is a measure that we call Nugent score, and I'll come back later to it, which is a research tool that's used for the diagnosis of bacterial vaginosis, and in this woman, the Nugent score was pretty much very high, indicative of bacterial vaginosis the whole time this study was going on. However, those women did not report symptoms related to bacterial vaginosis and did not seek medical attention. The third group of women had microbiota that changed constantly over time, very rapid changes, and you can see again the same color red orange here is lactobacillus innus and chryspatus, lot of changes in and out of a little bit of lactobacillus, no lactobacillus, but very diverse type of microbiota. So we were trying to understand what drive those changes, and we were able to model the change that we see over time by modeling this, what we call the rate of change of this index, the Jensen-Chenin divergence, and if you want it's an estimate of stability, and what you can see here is that during men's C, stability of the community is very low, indicated by this high index, high rate of change, and as you go towards the middle of the cycle, community stability is very much lower and stabilized, and then it comes back up and go up to being a little more unstable. If you map on top of this estrogen level, what you can see is that stability of the community is associated with higher level of estrogen throughout the menstrual cycle as well as higher level of progesterone. The factor that changed dramatically the community stability were obviously men's Cs, so the time in a menstrual cycle was very important, but as well as sexual activity to some extent, but not in all community background, so certain type of community dominated by different type of microbe tend to be more affected by sexual activity than others. So just to give you an idea of how dynamic those communities are, we just put this little movie together. So what you can see here is a heat map that shows time on the x-axis here and the different taxa and their abundance with red being the higher abundance here over time. And what you can see here is this ball here represent the state of the community where this black bar is, and the black bar is gonna move and the ball is gonna move into this vaginal space which I described earlier. And so remember, this is over 16 weeks in this particular woman and you can see how dynamic the community is. It's changing completely from being dominated by lactobacillus crispatus to not dominated by lactobacillus innards. It comes back here and it's moving a little bit away from even having lactobacillus and it's coming back and it's now gonna go all the way to this community that does not contain much lactobacillus. So what does this whole mean? So if you look carefully, if you map on top of this tetrahedron that I showed you, this PCOA plot, we can see that as I showed you earlier on this longitudinal pattern as well is that this community state type four here that we call is associated with high-nudgeon score. So potentially this tool that's used for diagnosing bacterial vaginosis as well as high pH. So what does it mean to have high pH and high-nudgeon score? So when we look at this, we see that in about 25% of those women in our studies are in those states that don't have lactobacillus, that are not dominated by lactobacillus. And those obviously have this high-nudgeon score and high pH, there's a strong association. So there's a pretty extensive literature that shows that when you have high-nudgeon score and to some extent pH, there's a strong association with an increased risk of sexually transmitted infection, acquisition and transmission, included HIV, as well risk and pregnancy from preterm birth. So we have women here in this particular study which are asymptomatic, apparently healthy, but potentially are at increased risk of sexually transmitted infection and other adverse outcome. So how can you summarize all this? I think we can summarize this by saying that throughout those period, throughout time, there are windows of higher risk to, for example, sexually transmitted infection that open and close on a temporal scale. And how can you model this and how can you represent it graphically? Think about this community state type four that's in the middle of this stitch right around that I showed you. Here what I did, I traced on this figure here the pattern of change of this particular community. Now think of every time the white line enter this sphere which represent that community state type four, this particular woman is at increased risk of, for example, acquiring sexually transmitted infection. Obviously it's only if she's exposed to it. And now we have this woman which I just showed you which conflictuate widely between different lactobacillus and enter this state only once compared to this one who was almost constantly into that state. And then we have that third woman which pretty much has not changed over time and stayed the white line just pretty much went over there. So what we have here is a state that gives higher risk to potential adverse outcome. And we wanted to know if this community stability and dynamics, so that's how we kind of like summarize it that the frequency and duration of this state actually might actually represent a better risk to disease. So the resilience of the community not just to lose, just to go into that state might actually represent the risk. So a low resilient community, a community that enter that state more often is at increased risk than others. So that's very interesting to know this but I think still this is a huge gap. While we know that we might be at high risk we still do not understand the underlying causes for this stability and this dynamic. And so we wanted to understand this at the molecular basis and look at a potential association between stability and this susceptibility. So we win this and we wanted to look at the correlation between the gene content of the community and that we have in different species as well as community stability. And to do this we perform a metagenomic analysis of a large number of, I think 50 different sample and we looked at comparing the genome of some of their member. We picked one of the member which we thought was very important and that member is lactobacillus innus. Lactobacillus innus is often associated with this state that we call bacterial vaginosis as well as often associated with frequent fluctuation. So by metagenomics we can actually very nicely generate compositional structure of the communities and you can see we have communities that have very low amount of lactobacillus innus, some that have high amount of lactobacillus innus and some that have lactobacillus innus with other type of lactobacillus. Some of those sample which are in black were sampled actually from community that already been characterized long internally and the sample in red were sample that were sampled cross-sectionally into certain patient. So the idea of sampling those one that come from community that already characterized long internally is that we might be able to associate a potential genotype of lactobacillus innus with community stability. So by comparing all the genome of all those lactobacillus innus which we were able to reconstruct from the metagenome we were able to construct this phylogenic trait and what this tree number one tells you is that lactobacillus innus is a very diverse species. It has a lot of different strain which genomically are somewhat different. Number two, we can now map onto all those different type of lactobacillus innus or strain of lactobacillus innus, those longitudinal profile. And when we do this, what we observe is that there are three branches on this tree that are associated with either instability. You can see the community going from innus dominated to lack of innus and so on over time. So those open and closed window of opportunity for infection or you have lactobacillus innus genome that are associated with communities that almost where lactobacillus innus cannot become the dominant member. And then we have branches on the tree that are associated with strain of lactobacillus innus that belong to community that are extremely stable over time. And this is even more important if you think and start adding on top of this a little more metadata that we collected at the same time and some of those sample which are lighted in color here and they shown here, those come from women who actually were diagnosed where came to the clinic because they have chlamydia. So now we have an association of certain type of lactobacillus innus associated with certain type of community that might provide a risk. And now we have those type of innus associated with the presence of chlamydia. So the drivers of community stability we see that there's different kind of lactobacillus innus genome that appears to be associated with community resilience to enter this community state type four. And we can ask the question is does this fragility, potential fragility of a single species in the community could drive the entire community to go into a despiotic state. And that state might carry risk to disease. So we now have this hypothesis that we are testing which we hypothesize that the vaginal microbiome its stability and its dynamic meaning the frequency of duration of this community state type four might be a better representation of the estimate of risk to infection and that this stability and instability can be driven by the lack of resilience or fragility if you want or increased fragility of certain keystone lactobacillus species into the community. So I just wanna finish quickly and I'm gonna highlight the different gap and challenges that I could come up with when it comes to studying dynamics and especially I guess issues related to women's health. I think that we haven't really solved this question of cause or effect of those changes that we see. So we really need to collect more sample prospectively into longitudinal study. I think that the time is really there. We have the technology, we have it's cheap enough that we can do it. We can do those large study where we have a lot of sample collected frequently and in a lot of subject. And so we need to better understand the driver of dynamic and stability or instability not just in disease but also in health because even in an healthy state we can be at risk and I think it's important to understand health a little better and do this not only in reproductive age woman for example in our case but also throughout a woman's lifespan. We need to understand what shifting microbiome means functionally. So looking at both the role of the microbiota but as well as the host. The host is still one in the vaginal microbiota is really largely unknown. So we need to evaluate this dynamic using different measure than just compositional measure. Metabolomics, transcriptomics both on the host and the microbe for transcriptomics as well as the immunological status of the host. We need to also expand our study to not just looking at the bacterial component of the vaginal microbiota but including phage and viruses as well as fungi. And all this needs to be used to develop predictive model of the stability and instability when a community is in what appears to be an healthy state. So and those model needs to account for our different health practices antibiotic use, hormonal replacement therapy as well as behavior. And all this needs to be used to translate this information in better diagnostic prognostic as well as treatment and moving the treatment of condition in women towards more of a personalized medicine. And I think that this association very strong association of certain strain of lactobacillus with women over time I think really calls for this more personalized medicine to the treatment. And with that I'd like to thank all my collaborators both at IGS here in Baltimore, Larry Fourney's group at the University of Valdeau and the clinical partners that are essential to do this kind of work. Was that the happy team? No, I want to answer a question. We have time for one question. Hopefully somebody will over here. Jacques, I think you covered it or I suspect that it's in there but I really appreciate your last slide except for the incredibly small font size but the question I have is how much of the gaps that you're talking about would also apply to other systems? Could you just sort of talk a little bit about I don't think they were just relevant to vaginal studies. Well I think that most of those gaps will deal with can be addressed in other and apply to other body site. I mean the lifespan, I mean Ruth I think made a good case for understanding this through the lifespan. Using different omics technology but not just looking at the microband or host, all this again applies to every body site, every disease that we can think of that's where the microbiota can have a role. And the translation, I mean this is all what we want. This is where we're going. This is why we're doing this kind of work. It's to take the finding that we do and transform them and to translate them into better diagnostic, better prognostic predictive tools as well as treatment. So I think they're all very translatable to any body site, any microbiota. I just have a quick question about other factors that you didn't mention like douching or a frequency of partners, the number of different partners or what, not just the number but what those partners are like, there must be all kinds of other factors that affect the vaginal flora. Absolutely and we have evidence now because we've been collecting sample prospectively that we have evidence that a single sexual act can actually introduce not overgrowth but introduce new microbes through semen and those microbes can actually take over the community temporarily and over a longer period of time. So no, you're entirely right. Douching is definitely an effect. I mean, we, Rebecca Brotman here published these interviews. Is it a good effect or a bad effect? Does it get rid of bacteria or does it introduce new bacteria, douching? I don't think douching introduced new bacteria. I guess it depends what product you use but it definitely has an effect on this dynamics and the amount of change that you'll see over time after douching. Thank you very much. Okay, thank you and let's thank Shaq one last time. This is okay. The last speaker for this morning's session is Rick Bushman, composition and dynamics of the human virum.