[00:00:05]
>> I'm excited to present all of you today it's definitely an honor and privilege to be here at Georgia Tech and I'm excited to meet many people who I have interacted with on line or have followed their work our read many of their papers and actually get to meet many of them here today and the students that have worked on the work with bam it's exciting to be able to present to you today I'm so yes I am actually Ford for sift I'm in the school of chemical engineering at Oklahoma State University and your very own Dr Mitchell is an alum of our program for undergraduate and so we're in Stillwater Oklahoma and it's a great place to live and work and so I hope by the end of it I have also give you a little promo about some open positions that if some of you might want to live and work there in the future there are opportunities so we do multi scale modeling of primarily tissues in the human body also treatments those are mostly chemical in nature but they can be physical in nature something like ultrasound we also look at toxicology in a lot of different senses so that could be chemical toxicants could be dysregulation hormone system but it also could be an infectious agent.

[00:01:16]
So I would be remiss if I did not acknowledge my team and our sponsors and so this picture is my team as of May We've had some variation since then but that version of the team included 2 graduate students I mean who is in the foreground she is the 1st ph d. student to finish from my lab she now works in a pharmaceutical company doing computational modeling Steve is a senior ph d. student Karley is now an investment in the rest of the teams undergraduates now I have a few more graduate students and I have openings for more and the work that I'm talking about today was primarily sponsored by the local Most Center for the advancement science technology and the Herald have diabetes center these are seed grants that lead to the innocent career ward and how relevant projects they are we also recently got this in i h grants and I'll explain more there but all the projects I'm talking about today were preliminary results to put into that and I insured and then there's been a number of undergraduate research sponsorships for scholar development in my lab different stages of my labs different students on the teams have been involved in publications which today I'll be showing sharing work that Stephen menu contributed to but also Anya laying ye Carly Jacqueline Michel.

[00:02:32]
Grace and Alex but then 2 collaborators So James crawlers and Harvard biology that they was at the time in Princeton aerospace engineering and so different collaborators at different times on the published work they all talk about it. So what do I mean by multi scale there's lots of different physical Times link scales and time scales involved in the human body and so there could be there is genetic information that's processed and translates into proteins and other chemicals that interact inside of single cells as molecular networks the cells give rise to cellular networks how multiple cells in a population interact and how heterogeneous populations of these cells interact each other and form complex networks that give rise to tissues then that those can interact for organ function or dysfunction and then that gives rise to the whole individual organism in this case a human body and then multiple humans with different characteristics and give rise to societal and social network types of behavior or population level phenomena my lab mostly focuses on cellular and tissue level so somewhere in between the organ and the molecular level for our types of the scales that we're bridging.

[00:03:50]
We also like to think about this paradigm where if you're on this level of where there's drug actions on targets there's great work and competition chemistry and biophysics on individual molecular interactions and a lot of work in Systems Biology about intra cellular networks these are reaction networks happening inside of an individual type of cell that drive how genetic information may be translated and then there's the tissue level how do multiple cells work together and actually give rise to functions that you can observe in tissue samples and in organisms and then the things you can typically measure and a patient are concentrations in their blood biopsy samples concentrations in their urine so we're very interested in connecting to these kind of clinical indicators and analyses about what's happening at the tissue level in a disease progression and we're also very interested in long scale types of processes or not just molecular interactions that may be on fractions of seconds but things that are chronic diseases that may be decades before you start to see damage or the diagnoses of damage so I'm going to tell you a few vignettes today.

[00:05:05]
I figured that this is a very broad audience and that you all had very different backgrounds and so I'm going to give you kind of a short commercial like stories about a few different areas in which we've published while I've been an assistant professor and I encourage questions in any of these areas once we finish so the 1st story I'll tell you is about immunology in him if there immunotherapy and some work we did related to tuberculosis.

[00:05:34]
In another work we looked at cancer metastasis and so we had in 2 publications on this most recently this year. And we have done a number of collaborative projects that include effects of Pirtle fertilizers and pesticides and so that led to 2 different bodies of work one single paper and one set of free papers and so I'll talk about that bit more later as well and then the thrust for a large part of my lab has focused on diabetic kidney disease so we have quite a few publications from my lab on diabetic kidney disease and some earlier work with some collaborators on kidney physiology in general so what is some of the common threads with this we're looking at processes that typically involve multiple types of cells different chemical interactions often physical interactions Typically there's some sort of transport involved with porous materials that are changing due to chemical reactions the chemical species can move through those materials but the materials themselves are heterogeneous and dynamic and were often a common thread literally between these different is extracellular matrix fibers the threads and they are typically being remodeled in different diseases they may be accumulating in a dysregulation way or they may be being broken apart in a way and so we're interested in those kind of levels of tissue damage due to having too much fiber formation or too little fiber formation or too much degradation and how that changes the micro environment in the tissue and how the cells and other chemicals interact in those systems so 1st the story on in these immunotherapy and in the immunology.

[00:07:18]
So just a little bit of background about tuberculosis is an infectious disease that affects many many people in the world and here the little red pill shaped things are the bacteria and your mean system or a patient's immune system is generally pretty good about responding to a bacterial infection and many immune cells will come and try to fight the bacteria and what your body will do is form a set of fibers around the cellular cluster to basically provide a fortress around and imprison the bacteria excuse and we call this a granuloma So there's fibers that are accumulating in a very controlled way around this cluster of cells and that happens in a very orchestrated manner and the bacteria are still alive so they are Jorma and but they're sequestered that's called latent tuberculosis and if a patient wasn't successful able to form these granulomas they would have active tuberculosis and so we're not so much interested when they're 1st exposed but we're interested in how one who has latent TB It's believed that many many people up to a 3rd of the world's population may have latent tuberculosis and so no obvious symptoms but has had an exposure where they may have some of these bacteria dormant in their body and what kind of process is Dr active transition from latent to active t.v. And so it's known that active t.v. can happen when this fibrous wall breaks down around the outside and so it's kind of a prison break the bacteria reemerge and you can have active tuberculosis and so we were interested in studying this problem to understand the dynamics of different perturbations particularly from HIV exposure other types of immune compromising events or aging that could lead to this progression to active TB a long time after the initial exposure.

[00:09:17]
And so we thought about some of the chemical processes that can break apart these fibers and have captured that in a mathematical model and so we were very particularly focused on this degradation and this degradation can be controlled by some signals from the bacteria themselves the bacteria are trying to actively get out and so that's where we focus with the private paper we the details of this are not critical but I have a Here is my one mathematical side but we're including the dynamics of the changes to this college and barrier and how that depends on some signals from the cells themselves we do explored a number of different things for the sake of time I'm just showing the results here where in HIV patients have.

[00:10:05]
Much smaller number of some of their immune cells t. cells in particular and so we modeled if we suppress the number of t. cells consistent with having a later h.p.v. exposure how the disease how the bacteria would actually be able to escape and how they would increase and so we had a stable initial point but then after we started to deplete the t. cell population we saw this dramatic rise in the bacteria as expected but we didn't program that it would be responsive to t. cell numbers reprogram but it would be responsive to the bacterial or the leakage would depend on the college and fibers and so that particular work gave us some preliminary modeling efforts in the immune system which we're now leveraging in other applications that have more rich data sets so cancer metastasis is a project that from personal.

[00:11:00]
Personal care about family members who have gone through metastasis and you know just knowing the patient prognosis is pretty grim for those with metastasis we sought to understand more about this process and again there are some fibers that are remodeled and so this cartoon is illustrating a primary tumor with a basement membrane that's already been ruptured and then these squiggly lines are representing different fibers in the my current Vironment So that would be the space between a tumor and the blood vessel and.

[00:11:33]
That there is some changes that happen to these fibers that facilitate the moment migration of cancer cells and so here we call these we have these little yellow Pacman type things these are Proteas this here him in pieces that break apart college and fibers and so they're kind of eating their way through this matrix and then the little bow ties locks this is so oxidase this is a inside that cross links other fibers and so it actually forms stronger bonds between the fibers and actually gives alignment to the fibers and that's easier for cells to migrate along those fibers and so this is just showing that the combination of these 2 chemical fields can influence the micro environment in a way that allows sort of fibers highways to be developed for cells migrate more easily and so we used 2 different modeling approaches one was deterministic and had one spatial dimension so a p.t.e. based modeling approach and then this with the cells were known we have formulated equation we just had how the cell population would respond but then we did another approach called.

[00:12:42]
Agent based modeling is actually a hybrid agent base where the chemicals are considered to be partial differential equations there are insufficient numbers and the cells though are discrete and the tumor itself can have different behaviors stochastically over time and so I'll go into those and results in a little bit more detail.

[00:13:04]
So all of this work was from a master student who was in my lab one years of undergrad one year semester soon. And so we're working on the validation parts of that now after she's graduated but this is work that was sort of generalizable for different types of tissue environments that different cancers exist in and so we were considering how these migrations be different if you had a.

[00:13:30]
Randomly aligned college and fiber network versus uniform college and fibers and if they were had different densities and so we have our tumor starts in the center we have the tumor cells start to move out from a central part and we could quantify different things related to their average displacement and how sort of collectively they moved versus how you know maybe a few individuals moved out and so what we could see is that the randomly distributed fibers these are heterogeneous prosody actually diminish the cancer invasion rate so when we had uniform materials the migration was much faster and that has ramifications for a lot of cell migration studies are done in environments where it's a single stiffness type of media and so it's important if if the biological system was heterogeneous to actually incorporate those things for more realistic quantification of the migration speeds we also looked at different starting concentrations so if you had different densities of this fibers matrix How would cells respond and so we found that if you had a fire at higher fiber const concentration there are more fibers that the cells can than if they're more realigned and they can actually get out more easily in those different types of materials and so did a number of analyses there.

[00:14:56]
On the pestle aside and fertilizer projects particularly pesticides we had I went to a a workshop sponsored. By a math bio center in n.c. state that was very focused on trying to connect biologists to mathematicians and I consider myself an applied mathematician and. It was people with data folks and people with data modeling skills and come together on cool projects and so there was a James Crow.

[00:15:28]
At Harvard and he said like I've got this great biological data said but I would like to have more understanding of what are some of the mechanisms driving these bees behavior when they're exposed to pesticides in this particular pesticide is now widely used and it's often in the formulation it's in the plant seeds and it grows up into the plant and so it was designed to not be harmful to be but to target other insects and what they're showing in their data is that a low dose low chronic dose of a farm of the pesticide can have certain effects on the health of the bees not their immediate like medical health but their interactions in their nest and sort of community behaviors that then could have long term effects over multiple generations so this is showing this kind of blob is the nest inside of a box so they have basically kind of the bee honeycomb inside of this box and these are not honeybees but they're essentially the equivalent of their food pots their brood where the babies are and with there's not really a big difference between are with the control group as expected is very very low almost undetectable dose is not much of a difference but this low level small dose had a big effect on how well that they traverse their nest and so we're very interested in why and how and so they also showed some facts of with this same size dose What's the change in their activity how mobile are these bees and how what's the change in the time that they spend nursing and the change in their average distance from the center of their net structure and so with this data this was kind of the compelling motivation to look at the system with a modeling approach and knowing that I look at pharmaceuticals and how cells respond then we could translate some of the ideas of how bees responds to pesticides to how cells respond to chemicals.

[00:17:26]
And so this is a movie of what he can capture and so let's do a little bit of animation here so. So each of the bees has a little barcode chip on their backs and so they've got an image analysis software program that my collaborator wrote that tracks you to be follows them over time and they just follow them in these box boxes that are the newest and they actually have a lot of these different nests in parallel that they have video capturing systems on and so they have quite a robust set of data collection and automated tools but the general idea is that they're on some nest and we're tracking individual bees within this box and each be can move and based on the behavior data literature they believed that there would be some attraction of the bees towards being on the nest and some attraction to based on where the other bees were where the food is and that kind of thing but then they'd also just have their own choices about movement so we call that the random walk and that there resultant actual velocity vector the angle that they would move out would depend on how much they were they cared about their own sort of migration patterns versus how much they cared about being attracted to the others on their nest and so that's what the result there's indicating we put our code on get held as an open source repository for others that are interested in the dynamics of bees so this work is completely agent based there are no chemical fields in the background we have a whole series of choices that are in the algorithm so I won't I'm not expecting you people to read that but there's a number of different rules that are part of this and if this is true with some probability then these are the steps that the bees go through so working through a couple of these.

[00:19:25]
So one of our sub models we call bump so if 2 bees are moving near each other if they touch there's probably some kind of think about billiard balls interacting they probably have some interaction and maybe that is already moving if it's going to switch to inactive like stop moving that might be different if it runs into a neighbor and decides to stop moving or if it just gets tired and sits down and so and likewise if it's stationary if someone runs into it it's probably more likely to start to move versus it waking up or because one thing to be active in spontaneously moving so that's what we're calling these probabilities of active inactive and this is active to inactive inactive to active and then I'll explain the other acronyms in a 2nd so location they know that if you're on the nest or if the bee is on the nest it's taking care of some of the biological functions of the nest and so it's probably more likely to stay on the nest given that it's working its doing things if it's off the nest it's probably more mobile is going to be there we head out to forage for food or it's interacting with its neighbors or heading back to the nest so that's one set of drivers and then the other would be its social interaction state so just spontaneously starting and stopping is one social interaction state where there is no neighbor and the other one is the socially modulating were to collide and there's a transition between them we also had more realistic ness that were based on the data and so which ones are occupied by food or young and the actual geometric patterns that those comprise.

[00:21:07]
We cram it tries this and I'm not going to go into all of the details but we used several of their data sets to parameterize the model and then use other datasets for validation and that's in the Science paper and. Trying to show some of these different effects that are different on and off the nest and what the actual value is where we put in for the probabilities of transitioning and these different rules for the beast behaviors we also then extended the work for generalize not just this specific dataset but what if you wanted to explore other effects of the nest environment so that's in our front here's an ecology evolution paper so the types of things that are in the model we have been moving around before exposure and then after exposure you can see that after exposure to the process there are a lot more lethargic there are some that are still moving around pretty well but there's just a lot slower behaviors and lot more sluggish essentially And so we're able to capture those types of things in the model.

[00:22:13]
This is artificial this particular simulation but showing that we can. Change behaviors based on this attraction string this is not what will really happen but if we set the attraction strength to they only want to go to the center of the nest or the center of their group of neighbors we would have the behavior on the right and then realistically it's probably some combination of I don't care where anyone is and being attracted to neighbors and so we tuned the dataset for.

[00:22:43]
The 1st model we also looked at the effects of having these are model results for the control and the images of the in I Am is the actual name of. The pesticide and we showed that if we only turn on one or 2 of our the the mob lading.

[00:23:02]
The mobility but also the attraction those 2 factors are not enough in isolation to describe the actual model output behavior and this is the experimental data and so we do see the same magnitude of decrease in activity if we turn on both of our factors and the right shape on the time on the nests and the distance to the center and so those types of things are captured in this model so again we can explore what if the next was indeed more crowded and what if it wasn't so having kind of these more mechanistic understanding for the different social interactions of the bee was very helpful to the biological team to be able to better understand their systems and they propose these mechanisms but we came up with a way to model that and work to make it more a simulation platform Ok so the last topic that I'll discuss is our area on diabetic kidney disease and so we had a one year seed grant a 3 year project and now and the career project associated with this and so a lot of.

[00:24:08]
Different teams in my lab have been working on this project for the entire time a bit on the state so why do we care about diabetic kidney disease. This is a chart of the causes of kidney failure and you can see that diabetes is the major contributor pressure is another major factor and high blood pressure and diabetes are co-morbidities and so sometimes those with high blood pressure are are more likely they are more likely to get diabetes and vice versa so the association with diabetes and other relevant cardiovascular disorders is really important for kidney failure and this kidney failure is complete no kidney function whatsoever and needing dialysis and a major major health concern and in Oklahoma diabetes is a one of the leading causes of death it's also very rampant in our very obese state and so it has a lot of public health implications for Oklahomans and I would be surprised if that was not also true for Georgians and so in particular diabetic kidney disease is a major complication of diabetes that happens in the kidneys and this is a cross-section of one human kidney in the human kidney there are many thousands of nephrons and you might have heard that term before and that's the word in a fraud for the Study of kidneys is related to nephron and this is a part of the now from this is the filtration section of the next one it's called the gloom aerialists I'm not going to ask you to pronounce that by the end but that is a term that my students have to learn on day one and this is this is a big blood vessel that comes in this is a set of branched Loopt capillaries that then collect afterwards as an exit ing blood vessel and so the blood that has not yet been filtered comes in through this afternoon or Tiriel the blood gets filtered through this collection of capillaries and then filtered clean blood exits through the f one arterial what happens in this capsule is that's where you're innocent.

[00:26:18]
Did and it goes downstream and later is collected to form urine and so this is the starting point for that liquid waste removal in the body and so what should happen is blood cells and macromolecule stay in the bloodstream but water salt ions other small molecules can be excluded and the body has a really precise control system for regulating blood pressure and this is known and then also with the slightly downstream zone re absorbing water and salt ions as needed and so it's really important for salt balance for protein concentrations in your body and all of that so you can use a really really important the way that they detect damage to this is by the presence of proteins in the urine so something has happened where proteins are not staying inside of these blood vessels but are instead being.

[00:27:10]
They're leaving through these capillaries and making it into the urine and so these capillaries are coded by a number of cells and so there's an inner layer there's a middle layer and the outer layer is primarily composed of these potent site cells and so they have put processes that stick together the color interdicted Tate and there's a lot of stuff about the health of these particular selves and some of these layers that comprise what we call the globe Marial or filtration barrier and how if that is damage that's when you're going to start to see protein in the urine we want to understand how glucose this this chemical that's in the body and we're supposed to have it how it's dysregulation can lead to long term damage in these post light cells and can actually disrupt this filtration barrier so it's chemical engineers we've got fluid mechanics in here we've got filtration separation processes but we also have an chemical reaction that works and that are happening.

[00:28:06]
This is looking at that capillary bundle but in a cross-sectional view and so each of these circles is a cross sectional slice of a blood vessel and what we have this is that entrance blood vessel that's the exit blood vessel and this is the capsule on the outside and so this slice right through this capillary bundle in the middle we have what are called cells of the ms n g m And it's kind of like the stuff that's in between all those blood vessels holding it together.

[00:28:38]
So that's giving some structural integrity to the loop of bunches of loops of capillaries and then there are cells on the outside the cells and then there's the protein is the green stuff in the center that's in the blood vessels when it's damaged the glucose can trigger some things with the inner processes here the misandry them to secrete more fiber So again there's that fibrous link that we are interested in understanding more and can actually grow and stretch that zone which stretches the post sites and the blood vessels the pota sites themselves can get chemically dis regulated by the presence of glucose in the other things that gets created and then some of the other layers can thicken and be damaged and so and they also are known to all interact with each other so if one of them is messed up then it affects the rate that these other cells are messed up and so we're interested in understanding how they communicate with each other how the chemical fields move and how this damage occurs and that all the adds to the presence of proteins particular albumen in the urine and that's what's happening in the diabetic kidney disease so it's known that this happens it's known that it starts like this but some of those details in between a really hard to measure experimentally because all of the cell types are very hard to culture it's very hard to culture with the 3 d. interactions and having this correct micro environment and it's very hard to measure these in the body because they're really deep inside of like nested within the kidney very very small and it's important though as they damage they can't reverse them and so competition modeling approaches we can connect to publish studies on some of the aspects that were isolated but then bring together the whole system and use indicators about.

[00:30:25]
Clinical health of the patients over a long time scales to do the overall model validation so we've been looking at a number of different chemical processes involved and how hyperglycemia can affect them so. Elves in these layers and so this is the glimmer of filtration where this pink part is the blood vessel we've got various molecules in the blood in the fuel cell layer the basement membrane and the proto sites the green parts of those put processes that stick together in normal glucose everything's fine only small molecules make it through the filtration barrier as increasing clue coast those put processes start to peel off we get this regulated health and then eventually some of the cells may die and some of the may fall off completely because they're not attached anymore they're actually stretched out they pop off and proteins can make it through so we want to be able to chemically mechanically describe these processes and so we've been taking a multi scale approach where we have built a series of.

[00:31:23]
Cellular biochemical that works that are mostly ordinary differential equation based for what's happening in sci fi to a single cell given some input what does the cells to create that can interact with its environment what are some of the transport processes happening between different cells and then how we get the crosstalk and the tissue structure changes and using the agent based or the hybrid agent based approaches.

[00:31:46]
And so some of the work that we've done has focused on these bio chemical reaction network models so the 1st one on the left was the whole body and something called the wren an angiotensin system this is a hormone system that regulates blood pressure and it's very well studied and there's one the most influential category of medicines to slow diabetic kidney disease and prevent kidney damage in particular are medicines for high blood pressure at regulation and so they were approved for cardiovascular health improvements but they had unexpected positive benefits in the kidneys and so that is a category pharmaceuticals called Ace Inhibitors And so this process is angiotensin engines of hormones and in the presence of God and then as a catalyzing agent it produces angiotensin one in the presence of injured hence and converting in Zion and other in time it produces angiotensin to answer since and to is a major factor in all of these damaging processes in the kidney so that's why we were interested in particular in modeling it and answered since into feedback feedback and regulates running production and so we've created a model that looked at data on high blood pressure medication that was in patients with hypertension and patients with hypertension and damage kidneys and this was a project that I had a couple of undergraduates on and this was whole body just understanding the pharmacokinetics of these processes so it had to look at the whole body looked at this relevant hormone system and it had this effect of the drug so if we're going to treat this this is the drug that's going to act in the kitty but this is what it's doing the whole body but it doesn't have any consideration of poker sites or the kidney and it doesn't have any glucose dependency we want to know how does this change if you have high glucose or different glucose dynamics so in the 2nd model we focus on pota sites and it's no.

[00:33:46]
And that in culture the post cites actually have a very different written into tense and system expression than the systemic and different processes and so this project focused on how tense it is produced and regulated in publicized but doesn't have a drug and it doesn't have the right in feedback mechanism so then our next project was to combine these effects and so we added glucose dependency through the green arrow so those actually were part of the process but the rate constants there depend on the glucose level we also have added the Brennan feedback systems of this red colored line here and also the pharmacokinetic effect by having the ace inhibitor that can selectively inhibit that ace in Zion and so that was a more robust model we can do the drug concentrations are the same for local and systemic because it's just how much of the drug is in the body.

[00:34:42]
But we see different things for the concentrations so for the systemic You have much smaller engine tensing concentrations but in the kidneys you can actually have this really really elevated concentration so even if your medication. Correctly suppresses how much angiotensin is in the whole body there's still can be a local potentially toxic level in the kidney cells and so when thinking about therapeutics Yes you can measure how much is and what's true but you also need to know if that's affectively lowering it sufficiently in the kidneys and so that's what this work is related to and there's different dynamics in the impaired kidney versus the normal keeping.

[00:35:24]
We also can add glucose dynamics so we just took this from the literature on different dynamics of patients across the day or a select patient that was representative non-diabetic and they represented diabetic about different cycles of glucose and we could see this is just the glucose dynamics that early time this is if we give the pharmaceutical there some zone that is dominated by the drug action where didn't really matter what their starting glucose was the drug affectively took care of lowering the interest rates and concentration and then there's some region where it hasn't completely restored but the differences between the drug glucose levels and started to make an impact on how fast one would recover and so those could lead to some differences between different patient groups we also been looking at some of the different geometrical anatomical.

[00:36:18]
Complexities of the transport processes here and so this is a blood vessel this is the zone in between in these little jagged things or are proto site cells how transport happens in sort of these complex or complex transport patterns and if this is only is going to be thickening How does that play a role and how can we get it to thicken chemically over time so we've been looking at those interactions with the Plasma the liquid part of the blood glucose and protein molecule transport and so we've been building frameworks in copy cell 3 d. So this is a very generic just a blob but it can represent either individual cells or regions of of tissue as discrete agents and then it can also represent these kind of blue green stuff or chemical fields in the background that we can show different layers of different chemicals that are also in the environment.

[00:37:09]
And so this is one of our 1st representations of. Misandry on the podium of the Camarilla race for membrane of sight cells and those edges between the cells that get really close to each other and that right now we haven't coupled the move the growth of the basement membrane in the growth of the ms Anjum But with this platform we can do those things we've also modeled what happens if we bring in a fluid flow from the edge and bring in this is just fluid We also then had if we had.

[00:37:43]
Our glucose was just coming in from this edge how it propagated through the environment and I updated this for another presentation recently I did put it in here we had these where they were scaled correctly where they were of the same magnitude on each one I just sort of picked this life but anyway the diabetic patient had a much higher glucose concentration to start off with and so you're Shapeways these look the same but the diabetic would have much larger magnitudes and the concentration group would be able to propagate through all of those tissue much more.

[00:38:12]
Damaging way for a normal patient and then also these were responded how the hormone system inside of the protocell perfect cells were stimulated the next step is to connect what gets stimulated in here to what is secreted and how that then feedback on the environment changing the diet of changing the thickness of the tissue and changing the transport processes in conjunction So we're just a couple of months into 5 year project on this work so.

[00:38:43]
So we're excited about the different ways we can couple these things. So that leads me to where do we have funding so this is the end of Sept project so I do have him 2 new graduate students on this project and then one student who's finishing up on and the state grant that was the preliminary funding for this work so the main idea is to use your computer tools to create basically a virtual microscope very deep into these styles into these kidneys in order to understand how the disease progresses and to be able to couple the entire kidney function due to the the girl Mary Alice some of the 2 real transport fluid dynamics processes but particularly looking at the crosstalk between these individual cells and then how that gets rises to some of the biomechanics So we also through that have developed a number of educational activities and so the one that I've been part of for the longest is called The Summer bridge program this is for our incoming college of engineering freshman and they are doing basically a 2 week camp and they get to do you.

[00:39:47]
3 design projects that are each 3 days long and they get to explore different majors to this process and so for chemical engineering we didn't have a session yet when I started and you know chemical engineering design could be reactors it could be distillation columns as Lisa for our local senior design is very distillation focused on but I wanted an application that students understood more intuitively and most people have taken some kind of medication and could think about the tradeoffs of taking it too frequently or the prices of a or the size of the pills they could have some judgment on what would make a good pick.

[00:40:28]
Medication from a customer standpoint are patients in point and so we talk about chemical and by medical engineering through this application of treating high blood pressure so we use the same medications I was just talking about and we talk about how they block the constriction of blood vessels that drives high blood pressure and they help the blood vessel to react relax if you have inhibited this molecule ace and we talk about the process where chemical engineering takes raw materials moves them around transforms them often chemically and to get higher value products we might be designing processes we might be designing products that we're generally interested in this framework in a biological system we in a disease we may have broken some of the transporter transformation systems but we get to control the raw materials the pharmaceuticals to then try to get back to our original products in this case the right level of the hormone in the body to have the right blood pressure and so they can think about some of these introductory design principles they can also think about a little bit of chemical kinetics and mass balances through some examples we do but we built a Matlab buoy which is an app for being able to simulate this and so they can put in those sizes and those frequencies have a drop down menu menu with 5 different medications and normal and kidney function and we assign the teams different combinations of these and they get to compare them talk to each other they get to think about differences between the different pharmaceuticals they get to think about the differences in the patient groups and what part of the mass balance might be different so what was the input the output the generation terms etc and so they get to play with this they get to design their own case they can compare different simulations this is only showing one day but they can press a button for 7 days in the get repeated doses and so they're trying to think about the target for where they want to hit but it's open ended so they're like do you want to really close to 100 percent they want to close to 70 percent to this is the level of reduction of this hormone of the body.

[00:42:28]
And so it doesn't do anything if you keep it at 100 percent of its original but they have to think about well do I want to disclose to 70 percent so it's just barely effective or do I want to get it as close to 0 and they can think about some of those different trade offs they can also put it at those frequency of $1.00 to $24.00 times a day which is ridiculous but they can see the effects of that and make some decisions about the frequency and also the dosing amount and so it helps them think about this is open ended there's not a right answer and we won't tell them if it's right or not because there's lots of different options and so we make them we have the design a power point presentation just the single slide I give them a template they insert their results they insert what they learned about it and kind of the broader implications of what they learned and so they get to say Well I think the purple one is better for this reason and they give their rationale and so maybe there probably is their 1st presentation of college because they're about to start you know a week later we count this for college credit but this activity we've used in a number of other formats 2 that are not this 3 day context we've used them for my kinetics cost and there's a women in science Expo which is shown here and that's actually the gooey on the table this is not a science museum it's a one day event in the state of Oklahoma that brings in $1000.00 middle school and high school girls and so we just have the booth at the table they can come up they can interact they said they want to design a pharmaceutical and they play with it for about 5 minutes they can see some impacts of that we also have to be simulator where we can say hey this is some of the different things that a pesticide might do to the bees and they can stew simulations but they'll say I learned about bees in my cost in the pesticides are harmful b.s. And so they're connecting to some things that they might be learning in middle school or high school but the fasting part we've also had students who are out for occasional technology centers who are like I've been doing programming language or using it in Fortran and so that has led to some really interesting discussions about the quality of the technology centers in Oklahoma.

[00:44:27]
Now we've also connected this to an event called grandparent University which is the run viral I'm nice to station this is an event that has grandparents and grandchildren come for a 3 day camp and so they get to major in something and so we are majors biomedical engineering and we are we do not have a b. in media partners we are chemical engineering department there's about 5 faculty in my department the do biomedical engineering work so we branded as biomedical engineering but we have one day on the kidney and one day on the long so my colleague does competition fluid dynamics and we have a number of games that we do with them too so we have the simulations but we also do some different physical demos of what it would like to be to be in a poor and to be a porous material we also have some activities with.

[00:45:12]
Experiments of filtration so we demonstrate filtration with water that then we can talk about what that's like in the kidney then we can talk about artificial kidneys and other vital technology innovations that are important we also use these kind of tools for numerical methods heat transfer kinds applications as well and undergrad at research So Claire and actually are both undergraduate women in my lab and these are undergraduates that were in the summer bridge program and these are of course elementary or the middle school high school children in the grammar of.

[00:45:43]
Women Science Conference. So the last little bit this project just started on September 1st and it's always one of those things that you don't know you're going to funding and then you suddenly have funding and you don't have students in place so I have one student on this project and I am looking for more and so students and or postdocs are very welcome and so this particular funding mechanism it's called and I each Mirah and it is a is funding the p.i. and the sort of research vision for the lab but it does not have specific names on the funding mechanism and so my lab is called Systems by medicine and pharmaceutics So this is quantitative systems by medicine and pharmacology but we're focused on multi scale tissue damage one of those processes the.

[00:46:27]
Involve extracellular matrix remodelling and in the proposal there's definitely an immunology focus so using our preliminary results in the 3 different main areas I told you about there were human biomedical we're looking at across spectrum of inflammatory disorders so things like arthritis information and bone remodelling due to inflammation and immunotherapy and so predominantly cancer immunotherapy is a really emerging new pharmaceutical technique very active area of f.d.a. approvals and so getting to work with cells of the immune system that interact with these remodelling micro environments and interact with cancer they also may interact with other tissues for other functions so leveraging some of the immunology work we've done and the metastasis projects and the remodeling projects in kind of one environment now this is from a review article that came out this year but a lot of the ideas that we put into our proposal we have active collaborators that look at bone remodelling experimentally and on can't you know therapies for respiratory diseases and so we have data available as these projects get started to work on.

[00:47:35]
And so with that there are media openings this is the current team and so I have 2 new ph d. students this sorry this fall but they were hired to work on the career project and so currently It was hired as a master student and like well the preliminary work you worked on as an undergrad just got funded so she's working on it now but I have opportunities for more people to join my team so even if you in the room are not interested tell your friends tell your networks so you have an ad you can reach me on my website my Twitter page or Twitter profile and by email and then also after the session Dr booth of Allah organize for a session q. and a just with students and post-docs and so if you have questions about any of these topics or if you are interested in post-doctoral opportunities I'm happy to chat and so that's the area that I have for new students to work on my team with fact I have just a summary slide and would welcome any questions from you all thanks.

[00:48:52]
Yes Cathy. Yes So my framers her predominately mechanistic 1st principles where we look at bio chemical reaction networks and transport theories the physics and chemistry behind processes and we're starting to add some biomechanics So we haven't done more the bioinformatics and data analysis techniques that your lab focuses on and but there's certainly room for doing that particularly on.

[00:49:32]
Understanding more of the the data that's out there that's already been collected so some of the text mining kind of work that you do to give to be able to find more data that we didn't know existed in might have different key words associated with it or to infer certain things we haven't done a lot with.

[00:49:50]
Statistical parts because we're not looking at necessarily the reaction networks inside of cells that we know are very hard to probe because we're looking at we know this particular one is known to be secreted and have an effect on the other one so if we were going to focus on interest cellular networks so we would have.

[00:50:08]
Looked a lot more into graph theory and things like that but I think there's a lot of great people including some in the room that are doing a lot of great systems biology work at single cell level and so relying on that area of expertise to inform some of the stuff that we do but no we haven't we are mostly focused on.

[00:50:24]
Connect the transport based model. Yes So depending on the application we choose the software that makes the most sense the time so if we are not doing fluid dynamics we're probably not going to use fluid software but we have done a project that I'm not showing here that has kept actual fluid dynamics in it because there's convection happening blood flowing and so then we use console for that one for our metastasis project we used Matlab for the p.d. model we used Compu self 3 d. for the hybrid agent and p.d. base model for the bees we use Matlab we're able actually to write her agent based models using vectors and matrices in Matlab because we only had about 100 G.'s and the largest system that they could put in the side a single nest would be on the order of a few 100 bees so it worked very quickly in matlab the other projects some Python some Matlab we also sub self in Fortran.

[00:51:34]
What else that's predominantly is just. Our. Room. And safety and. You. Know it's a very. It is great for a lot of reasons but you have to be able to have computers or apps to do it but we in our techniques having with the grandparents grandchildren having the kids help the grandparents install them out of apps was very helpful piece because the kids were really savvy with it we also thought about other technology pieces that weren't necessarily simulations so you can buy these t. shirts that have essentially secure our code but it looks like the human body and you do your apps on it and it looks like a virtual reality version of looking inside of the body it's not that person's body but it kind of can be perceived as it may be your body because you're wearing this t. shirt on your body and so we use those these apps based educational tools that really hook them and they were like my gosh there's a kidney and that's connected to this and they could turn on and off different layers and so even if it's just a visualization of something that's real and connects to them even if it's not all of what you want to do and not a simulation of how this disease progresses and all of that it's certainly a technology that way to get them engaged in a way that they couldn't dissect a human and or even an animal would be expensive and there'd be safety protocols but having this very visual virtual reality experience was really cool and we let them take home the teacher like $5.00 and the app was free but that was such an important part of that program for those participants we have had the criticize President that a lot of my apps happen to be in matlab and so if I'm going to give this to a high school teacher they don't necessarily have a Matlab license but we now Matlab has the ability to package your apps as a complete standalone where it's just an executable file and they can then download it and have all of the infrastructure they need without the simulation Vironment I know console is moving that way Python apps are certainly that way already other tools are shiny to be hosted on the websites and so.

[00:54:05]
Chris case leaks and I and some others recently just submitted a paper on some different app environments that chemical and. Faculty and investigators can use to teach students but also the public about different technologies using kind of these and capsulated software programs that then you can distribute so using I.O.'s.

[00:54:24]
Consul Excel Python are as great rule ways to have kind of a visual stimulation environment a web him over molecular dynamics was another one we reviewed where they can give some input they see some output it's some subsection of what you may have in your actual research model but they can understand some process through that and thinking about ways that actually make it visually stimulating So I showed you the really early version of my b. app but the more recent one we put a picture of the be at the top so that then they walk by They're like what's this thing with bees rather than there's blue circles moving on the screen that it actually connects to them so yeah yes.

[00:55:11]
Yes Yes So great question so we for the cell migration we looked at a few experimental data papers that had a they have measured cells over time on a certain material that we considered to be sort of our average material and then we extended the ranges beyond that we also it's a money Carlo simulation so there's not an explicit time step so we looked at there was another paper that used the same software platform that also looked extracellular matrix but they didn't have all the same chemical environment processes that we did and so they had a technique for assessing basically figuring out based on the migration of a cell how that bin back calculated to the distance it went through so he knew how much distance we were representing but we didn't actually know how much time corresponded to each Monte-Carlo step because it just kind of.

[00:56:00]
Advances it doesn't say it's one second of physical time so we were able to get that the fusion coefficients we look at. For a range of different projects. We often look at the infinite diffusion coefficients that are measured and then we'll look at different hydrogen and theories about how you would move through different hydrogen or through random fiber networks based on the fibro city or the prosody of those materials and so in that particular one it was a bit more simplistic than that because we were comparing to existing math models that didn't have a lot of the remodelling effects and so we assume some of the same parameters as those in the literature but we have much more sophisticated ways to take the diffusion into account.

[00:56:59]
Yeah that's a great question so those are things I have proposed but not have done so most of the simulations we do right now Ok so not cancer project the other 2 the simulations that we have done are fast and so there hasn't been a big burden My seems to them on their laptops when we add the company 3 d. though with this stochastic agents that move progress over time we're doing those right now on a dedicated desktop machine takes depending on the length of time we want to actually simulate minutes or hours to do that we know for long term projects they're going to need to be bigger than that but we're in prototype development stage right now we there is a parallel version of company that can run on our universities high performance computing center cluster and so we haven't gotten to the point yet that we've needed that but we know it exists and so for our production scale simulations and we'll be using that so that we can take advantage of the parallelism and so.

[00:57:57]
Some of the processes may become of complicated in the the multi scale part but right now they've been working independently so. But short term independently so thank you.