Cadmore media player playing video Systems biology with Andrea Califano
Video Player
Accessibility
Using forms mode, you can utilize keyboard shortcuts to make usability more efficient.
Space bar will play and pause the video. Pressing escape will close all displays and pressing it again will jump to the end of the player.
You can press A to open the Keyboard Shortcut Dialog which also provides an audio description of the player and all of its features.
Language
1x
Play Speed
Adjust Volume
Transcript
Search
Result Count
of
Click to jump to result
Search
Re-center
Segments
Result Count
of
Click to jump to result
Search
Name:
Systems biology with Andrea Califano
Description:
Systems biology with Andrea Califano
Thumbnail URL:
https://cadmoremediastorage.blob.core.windows.net/dff5c082-4352-4753-8469-c4083e0cfe86/videoscrubberimages/Scrubber_15.jpg
Duration:
T00H16M01S
Embed URL:
https://stream.cadmore.media/player/dff5c082-4352-4753-8469-c4083e0cfe86
Content URL:
https://cadmoreoriginalmedia.blob.core.windows.net/dff5c082-4352-4753-8469-c4083e0cfe86/BTN-2019-0003.mp4?sv=2019-02-02&sr=c&sig=vouozRrG2HVDnWSX0uyKD3nMwGA4QywApYm7L%2FMlYg8%3D&st=2025-06-30T21%3A52%3A33Z&se=2025-06-30T23%3A57%3A33Z&sp=r
Upload Date:
2020-01-13T00:00:00.0000000
Transcript:
Language: EN.
Segment:1 An introduction to systems biology.
[MUSIC PLAYING]
Segment:2 What is systems biology?.
ANDREA CALIFANO: With the initial discovery of the oncoviruses and oncoproteins more than 30 years ago at this point, we imagine that cancer could be solved in a relatively straightforward way. There was a broken-- there was a lesion in the genome that was induced by a variety of different causes, and therefore out of that lesion would come this aberrant phenotype of the cell. Once we got past that stage, we realized that that was actually not all of it, and that cells can actually do all sorts of amazing things.
ANDREA CALIFANO: They can reprogram themselves. For instance, all of the cells in our bodies have exactly the same genome, and there's nothing more different than a liver cell and a brain cell. And therefore, cancer cells, even though they may have the same aberrant genome, they can be in all sorts of different states. In addition, almost every single cell in a tumor, in a tumor mass, has actually slightly different genetic mutations because they keep accumulating.
ANDREA CALIFANO: And so starting from a concept that seemed at the time very simple, we now have migrated to sort of understanding cancer as an extremely complex living organism. So now there's two ways you can go about it. You can build an inventory of all the parts that are broken in that organism one cell at a time. And what you'll find that way is that if you actually try to count the number of possible patterns of mutations that can give rise to cancer, that number is actually larger than the number of atoms in the universe.
ANDREA CALIFANO: So there's only-- only-- there's 10 to the 4-- there's 10 to the 80 atoms in the universe. There's about 10 to the 400 possible patterns of mutation, even if you only consider about 1,000 cancer genes. So this obviously says that you, at best, if you look at mutation one by one, you can find the very, very tip of the iceberg. It isn't even going to be 1/10 of the total mass, like, for water and ice.
ANDREA CALIFANO: It's actually going to be basically 1 quadrillionth of what the total number of patterns that could be. So the other way you can go about it is to try to understand how cancer works. Cancer works as a complex sort of set of levers and pulleys and gears that work together to make something happen. And so what we do in that case, and this is sort of the foundation of system biology, is instead of using brute force to study things one genome at a time or one protein at a time or one metabolite at time, we'd first try to build a model of how things work, for instance, how certain proteins regulate the expression of certain genes, or how proteins interact in a signal transaction cascade, until that model can allow us to ask very complicated questions in a very simple way.
ANDREA CALIFANO: Right? So it's the same way as if-- think about before Newton's laws of dynamics, if you wanted to know how a ball falls down an incline, you would have to measure the ball at multiple different points. Today, because we have the laws of motion, you don't even have to do that. You can actually compute it with a model that will tell you exactly what the speed of that ball is going to be and where it's going to get, even if the incline actually has turns and valleys, et cetera. So the idea is to turn biology from a empirical science into a very quantitative and predictive science.
Segment:3 What research is happening in your lab? .
ANDREA CALIFANO: So my lab has been the first to create comprehensive maps of regulation of human cells. And so, like, around the year 2000, we started to take cells that were associated with lymphoma, or with normal B cells in biology, and studying how every single gene was being regulated by protein called transcription factors.
ANDREA CALIFANO: And you can immediately see how these can become very complex, because there is about 2,000 transcription factor and co-factors that regulate expression of genes, that is 20,000 genes approximately. And so the number of possible interaction is 2,000 times 20,000 potentially, which is a very, very large number. And so-- and the more important thing is that in every single tissue, lineage, or even single cancer cell, the way these genes are regulated by specific proteins is actually-- can be actually different.
ANDREA CALIFANO: So, for instance, the way the classical oncogene like Myc, protein oncogene like Myc, is regulating its target in lymphoma cells can be dramatically different from the way it's regulating its target in a neuroblastoma cell. So this required using leveraging a branch of mathematical information theory. And we used an algorithm called Arachne. And now we have maps of regulation for virtually every human tissue. In fact, we can get maps of regulation that are specific to a particular individual by leveraging single cell biology.
ANDREA CALIFANO: Now that we have this map, the question is, what can we-- what can we use these maps for? I mean, it's great to have a list of all possible interactions, but what we really want to know is, what is this drug doing in this particular cancer cell? What are the key driver proteins that we need to shut down?
ANDREA CALIFANO: And so we built additional algorithms-- a key one is called VIPER, which stands for Virtual Proteomics-- which can actually take-- it works like Netflix prediction. So it basically uses the genes that are regulated by particular proteins as a proxy to figure out whether this protein is activated or inactivated. And so you can transform the gene expression signature of a tumor or of a single cell into a very precise, very accurate measurement of the activity of about 6,000 proteins, all the regulatory proteins.
ANDREA CALIFANO: And that does something quite cool, which is, first of all, we can ask, what are the proteins that truly mechanistically control the state of the cancer cell? Because those are the ones that are going to be the major dependencies. You shut them down, the tumor is going to not be able to stay in the state anymore. And on top of that, you can actually ask, what does a drug do to the activity of the entire set of proteins that we can monitor, right?
ANDREA CALIFANO: So normally we think of a kinase inhibitor, like a MAP kinase inhibitor, like MEK inhibitor, as targeting just MEK, right? But every single kinase inhibitor is actually not only inhibits a-- or activates a variety of other proteins through MEK, but it would actually inhibit, as a off-target effect, maybe another 40 or 50 kinases and god knows how many other proteins that are not even kinases. So what we can do with this approach is really rediscover pharmacology from scratch.
ANDREA CALIFANO: We basically put drugs in a cell experimentally, and we can generate gene expression profiles. And now we can figure out exactly which proteins increase in activity, which proteins decrease in activity. So if you know that a particular protein is a dependency of a cancer cell, you can now ask, did this drug shut it down, or does this drug didn't do nothing to that protein? And you can ask it not just for one protein. You can add an entire program that controls the state of the cell. And now you have a lot of proteins that are up in the tumor, and you're asking, is there a drug that will shut them down completely? OK?
Segment:4 What is OncoTreat?.
ANDREA CALIFANO: And so we develop this methodology called OncoTreat, which is the very first New York state CLEA approved methodology that is based only on RNA to predict which drugs will benefit which patients.
ANDREA CALIFANO: And so we have now been using this methodology very extensively in clinical trials to make a notice of conflict of interest. Because this algorithm will be licensed to a company that I started called Darwin Health, and so both I and Colombia are equity holder in the company, and so I need to make sure that this is known.
ANDREA CALIFANO: But the idea is that now we have response rates in patients, for instance, that have been transplanted into mice that were considered untreatable because they failed three to seven lines of therapy of greater than 60% where normally you would expect less than 5%. And these are, for instance, if you just do a negative control with drugs that were not prioritized by our approach, the response was actually zero. We didn't get a single response in these mice that were transplanted. So it's exciting. We have now started about six clinical trials based on this technology. Two have already closed, and we are now basically going through the results, which look very exciting.
Segment:5 What could be the applications in precision medicine beyond oncology?.
ANDREA CALIFANO: We are doing a lot of different studies that actually are not cancer related. Cancer is a great area for developing this type of methodology because the amount of data that is available is enormous compared to other fields. First instance, for every profile that you have in neurobiology, you may have 100 profiles in cancer.
ANDREA CALIFANO: And this is because cancer cells can be very easily cultured. They're very easy to access from patient biopsies, et cetera. I try, but it's been very hard to convince my postdocs to donate parts of their brains without having, you know-- So you end up having an enormous amount of material, data, reagents that you can use. But there is absolutely nothing that is specific to cancer in the methodology that we've developed.
ANDREA CALIFANO: And in fact, so for instance, we have applied them to Alzheimer's, Parkinson's, alcohol addiction, diabetes. The latest study in diabetes is how to block reprogramming of the beta cells in the pancreatic islets into alpha cells, so that you can actually prevent the onset of diabetes. So there's quite a number of approaches. And the nice thing is that they all-- they all are based on these very fundamental, if you want, maybe sort of a new theory of disease, which basically says, how is it possible that you end up getting the same disease if you actually have this incredible variety of mutation or variants that can give you that disease, right?
ANDREA CALIFANO: So you have potentially 1,000 genes that can change your-- that can induce obesity or that can induce diabetes, right? How is it possible that the disease actually looks so much the same even though the mutations are so many? And so the theory that we've formulated is that in fact, one way to explain this is that the mutations or the variance that you have are all upstream you know almost like a hour glass model of a very small number of proteins that we call master regulators.
ANDREA CALIFANO: And then these proteins are the ones that introduce, that implement the specific programs that are necessary for the cell to be in a disease state. And so the idea is that instead of actually acting upstream here at the mutational level, because there's an enormous number of different varieties, if you act at these bottlenecks, there are very few proteins that we actually have to target. And this is exactly how OncoTreat works, the same way in which you could work-- for instance, we're now doing something equivalent to what we've done in cancer or diabetes.
ANDREA CALIFANO: We're identifying drugs that can completely reverse the signature, by which these beta cell reprogramming to alpha cells. And that could then be used as drugs in precision oncology.
ANDREA CALIFANO: Why do we call it precision? We call it precision because-- in fact, this is a term that somewhat offends oncologists because they say, I've always been doing things in a precise way. I've always developed treatment plans that are completely specialized to the individual. So what is new?
ANDREA CALIFANO: What is new is that we are using really high-dimensional data, very complex models, very quantitative models, to make a decision, to make predictions. And we then use very systematic approaches to validate these predictions, preclinically and then clinically, to the point that, very much like the model of the equation of motion that Newton developed, you can now have the equation of cancer, or the equation of diabetes, that you can use to predict whether a particular patient will respond or not respond to a particular drug.
ANDREA CALIFANO: And if they don't respond, how do we sensitize them by using, say, a second drug. And so this is-- the concept of precision medicine is really driving medicine from a more empirical base of trial and error, if you want, driven by certain principle, but relatively-- alludes to something where you have a very, very sophisticated model that is highly predictive.
Segment:6 Do you have an example of the application in precision medicine?.
ANDREA CALIFANO: We think of precision medicine in cancer as targeting a particular protein that is mutated, like an EGF receptor, with a drug called erlotinib or afatinib. And yet the idea is that it sounds great in principle because you say, I have a mutation. I have a drug that targets that mutation. What's more precise than that? The problem is that only about 35% of the patients that have the EGFR mutation, activated mutation, will respond to these drugs, and the vast majority of those will eventually relapse.
ANDREA CALIFANO: Same thing in HER2-positive breast cancer treated with trastuzumab. You got a mutation in a receptor kinase protein. You target it with an antibody. You shut down completely that path. And yet about 30% of the patients don't respond initially and 70% will eventually fail to respond even if they did respond initially. So that's where there's a tremendous need for additional precision.
Segment:7 Do you have anything else to add?.
ANDREA CALIFANO: For a long time, there's been sort of a division between a group of investigators that were called computational biologists and a group that were sort of experimental biologists. And there was a time when I started first in the Computational Biology Group at IBM where we were starved for data. We literally had to go beg for getting data.
ANDREA CALIFANO: We are now at the exact opposite. The experimental biologists are coming to the computational biologists to ask them to process their data because it no longer fits in an Excel table where they can actually see it. But more importantly, I think what's happening-- this is certainly my case. We are starting to develop labs that are fully integrated. So there are labs where we do both the experimental part and the computational part.
ANDREA CALIFANO: And this, I think, is a extraordinary way to simplify the process because you generate the data that you need to create a model. You create the model. That model makes predictions. You can now put those predictions back into the lab and validate them. And from the validation, you can now learn how to improve the model and make it more sophisticated, more accurate.
ANDREA CALIFANO: And so at the end of this process, I hope that we will not think of ourselves as either being experimental or computational biology but rather as just being biologists, or oncologists, or whatever discipline of life science you're interested in.
Segment:8 What are your AACR 2019 highlights?.
ANDREA CALIFANO: It's extraordinarily difficult to talk about AACR highlights because this is such a humongous conference. You know, there's 20,000 people. There's millions of things that are going on. Certainly an area that is becoming of great interest is the use of immuno-oncology approaches. And there is a tremendous spurt of development. I think also there's been a bit of a slowdown in terms of sort of thinking that the mutational space will solve the problem of cancer. And so this has really created an opportunity for these orthogonal approaches based on systems biology and other types of discipline, I think.
ANDREA CALIFANO: AACR likes to call it convergence. I think Phil Sharp come up with that term. And basically sort of bringing together a variety of different disciplines. For instance, before I was talking about information theory, which comes from a branch of mathematics, and biology to come up with models that can actually be predictive of cancer response to certain drugs.
ANDREA CALIFANO: I think this is a very important process that has to go on. Hopefully, the AACR will continue to support that. Another key area, I think, is going to be around doing a better job with models. How do we model cancer so that when we actually go to a clinical study our probability of success increases quite substantially? And this can only be done if more sort of high-fidelity models are used, and also if we really learn to understand what a model is good at.
ANDREA CALIFANO: So for instance, tomorrow in my presentation, I will discuss how we used a cell line that is actually histologically not even from the same type of the tumor that we were studying, which ended up being the most faithful model we could use to actually predict response to drugs in patients that had been validated in actual explants from patients with that particular tumor type. So I think this is, again, an area that is expanding quite dramatically.
ANDREA CALIFANO: [MUSIC PLAYING]
Segment:1 An introduction to systems biology.
[MUSIC PLAYING]
Segment:2 What is systems biology?.
ANDREA CALIFANO: With the initial discovery of the oncoviruses and oncoproteins more than 30 years ago at this point, we imagine that cancer could be solved in a relatively straightforward way. There was a broken-- there was a lesion in the genome that was induced by a variety of different causes, and therefore out of that lesion would come this aberrant phenotype of the cell. Once we got past that stage, we realized that that was actually not all of it, and that cells can actually do all sorts of amazing things.
ANDREA CALIFANO: They can reprogram themselves. For instance, all of the cells in our bodies have exactly the same genome, and there's nothing more different than a liver cell and a brain cell. And therefore, cancer cells, even though they may have the same aberrant genome, they can be in all sorts of different states. In addition, almost every single cell in a tumor, in a tumor mass, has actually slightly different genetic mutations because they keep accumulating.
ANDREA CALIFANO: And so starting from a concept that seemed at the time very simple, we now have migrated to sort of understanding cancer as an extremely complex living organism. So now there's two ways you can go about it. You can build an inventory of all the parts that are broken in that organism one cell at a time. And what you'll find that way is that if you actually try to count the number of possible patterns of mutations that can give rise to cancer, that number is actually larger than the number of atoms in the universe.
ANDREA CALIFANO: So there's only-- only-- there's 10 to the 4-- there's 10 to the 80 atoms in the universe. There's about 10 to the 400 possible patterns of mutation, even if you only consider about 1,000 cancer genes. So this obviously says that you, at best, if you look at mutation one by one, you can find the very, very tip of the iceberg. It isn't even going to be 1/10 of the total mass, like, for water and ice.
ANDREA CALIFANO: It's actually going to be basically 1 quadrillionth of what the total number of patterns that could be. So the other way you can go about it is to try to understand how cancer works. Cancer works as a complex sort of set of levers and pulleys and gears that work together to make something happen. And so what we do in that case, and this is sort of the foundation of system biology, is instead of using brute force to study things one genome at a time or one protein at a time or one metabolite at time, we'd first try to build a model of how things work, for instance, how certain proteins regulate the expression of certain genes, or how proteins interact in a signal transaction cascade, until that model can allow us to ask very complicated questions in a very simple way.
ANDREA CALIFANO: Right? So it's the same way as if-- think about before Newton's laws of dynamics, if you wanted to know how a ball falls down an incline, you would have to measure the ball at multiple different points. Today, because we have the laws of motion, you don't even have to do that. You can actually compute it with a model that will tell you exactly what the speed of that ball is going to be and where it's going to get, even if the incline actually has turns and valleys, et cetera. So the idea is to turn biology from a empirical science into a very quantitative and predictive science.
Segment:3 What research is happening in your lab? .
ANDREA CALIFANO: So my lab has been the first to create comprehensive maps of regulation of human cells. And so, like, around the year 2000, we started to take cells that were associated with lymphoma, or with normal B cells in biology, and studying how every single gene was being regulated by protein called transcription factors.
ANDREA CALIFANO: And you can immediately see how these can become very complex, because there is about 2,000 transcription factor and co-factors that regulate expression of genes, that is 20,000 genes approximately. And so the number of possible interaction is 2,000 times 20,000 potentially, which is a very, very large number. And so-- and the more important thing is that in every single tissue, lineage, or even single cancer cell, the way these genes are regulated by specific proteins is actually-- can be actually different.
ANDREA CALIFANO: So, for instance, the way the classical oncogene like Myc, protein oncogene like Myc, is regulating its target in lymphoma cells can be dramatically different from the way it's regulating its target in a neuroblastoma cell. So this required using leveraging a branch of mathematical information theory. And we used an algorithm called Arachne. And now we have maps of regulation for virtually every human tissue. In fact, we can get maps of regulation that are specific to a particular individual by leveraging single cell biology.
ANDREA CALIFANO: Now that we have this map, the question is, what can we-- what can we use these maps for? I mean, it's great to have a list of all possible interactions, but what we really want to know is, what is this drug doing in this particular cancer cell? What are the key driver proteins that we need to shut down?
ANDREA CALIFANO: And so we built additional algorithms-- a key one is called VIPER, which stands for Virtual Proteomics-- which can actually take-- it works like Netflix prediction. So it basically uses the genes that are regulated by particular proteins as a proxy to figure out whether this protein is activated or inactivated. And so you can transform the gene expression signature of a tumor or of a single cell into a very precise, very accurate measurement of the activity of about 6,000 proteins, all the regulatory proteins.
ANDREA CALIFANO: And that does something quite cool, which is, first of all, we can ask, what are the proteins that truly mechanistically control the state of the cancer cell? Because those are the ones that are going to be the major dependencies. You shut them down, the tumor is going to not be able to stay in the state anymore. And on top of that, you can actually ask, what does a drug do to the activity of the entire set of proteins that we can monitor, right?
ANDREA CALIFANO: So normally we think of a kinase inhibitor, like a MAP kinase inhibitor, like MEK inhibitor, as targeting just MEK, right? But every single kinase inhibitor is actually not only inhibits a-- or activates a variety of other proteins through MEK, but it would actually inhibit, as a off-target effect, maybe another 40 or 50 kinases and god knows how many other proteins that are not even kinases. So what we can do with this approach is really rediscover pharmacology from scratch.
ANDREA CALIFANO: We basically put drugs in a cell experimentally, and we can generate gene expression profiles. And now we can figure out exactly which proteins increase in activity, which proteins decrease in activity. So if you know that a particular protein is a dependency of a cancer cell, you can now ask, did this drug shut it down, or does this drug didn't do nothing to that protein? And you can ask it not just for one protein. You can add an entire program that controls the state of the cell. And now you have a lot of proteins that are up in the tumor, and you're asking, is there a drug that will shut them down completely? OK?
Segment:4 What is OncoTreat?.
ANDREA CALIFANO: And so we develop this methodology called OncoTreat, which is the very first New York state CLEA approved methodology that is based only on RNA to predict which drugs will benefit which patients.
ANDREA CALIFANO: And so we have now been using this methodology very extensively in clinical trials to make a notice of conflict of interest. Because this algorithm will be licensed to a company that I started called Darwin Health, and so both I and Colombia are equity holder in the company, and so I need to make sure that this is known.
ANDREA CALIFANO: But the idea is that now we have response rates in patients, for instance, that have been transplanted into mice that were considered untreatable because they failed three to seven lines of therapy of greater than 60% where normally you would expect less than 5%. And these are, for instance, if you just do a negative control with drugs that were not prioritized by our approach, the response was actually zero. We didn't get a single response in these mice that were transplanted. So it's exciting. We have now started about six clinical trials based on this technology. Two have already closed, and we are now basically going through the results, which look very exciting.
Segment:5 What could be the applications in precision medicine beyond oncology?.
ANDREA CALIFANO: We are doing a lot of different studies that actually are not cancer related. Cancer is a great area for developing this type of methodology because the amount of data that is available is enormous compared to other fields. First instance, for every profile that you have in neurobiology, you may have 100 profiles in cancer.
ANDREA CALIFANO: And this is because cancer cells can be very easily cultured. They're very easy to access from patient biopsies, et cetera. I try, but it's been very hard to convince my postdocs to donate parts of their brains without having, you know-- So you end up having an enormous amount of material, data, reagents that you can use. But there is absolutely nothing that is specific to cancer in the methodology that we've developed.
ANDREA CALIFANO: And in fact, so for instance, we have applied them to Alzheimer's, Parkinson's, alcohol addiction, diabetes. The latest study in diabetes is how to block reprogramming of the beta cells in the pancreatic islets into alpha cells, so that you can actually prevent the onset of diabetes. So there's quite a number of approaches. And the nice thing is that they all-- they all are based on these very fundamental, if you want, maybe sort of a new theory of disease, which basically says, how is it possible that you end up getting the same disease if you actually have this incredible variety of mutation or variants that can give you that disease, right?
ANDREA CALIFANO: So you have potentially 1,000 genes that can change your-- that can induce obesity or that can induce diabetes, right? How is it possible that the disease actually looks so much the same even though the mutations are so many? And so the theory that we've formulated is that in fact, one way to explain this is that the mutations or the variance that you have are all upstream you know almost like a hour glass model of a very small number of proteins that we call master regulators.
ANDREA CALIFANO: And then these proteins are the ones that introduce, that implement the specific programs that are necessary for the cell to be in a disease state. And so the idea is that instead of actually acting upstream here at the mutational level, because there's an enormous number of different varieties, if you act at these bottlenecks, there are very few proteins that we actually have to target. And this is exactly how OncoTreat works, the same way in which you could work-- for instance, we're now doing something equivalent to what we've done in cancer or diabetes.
ANDREA CALIFANO: We're identifying drugs that can completely reverse the signature, by which these beta cell reprogramming to alpha cells. And that could then be used as drugs in precision oncology.
ANDREA CALIFANO: Why do we call it precision? We call it precision because-- in fact, this is a term that somewhat offends oncologists because they say, I've always been doing things in a precise way. I've always developed treatment plans that are completely specialized to the individual. So what is new?
ANDREA CALIFANO: What is new is that we are using really high-dimensional data, very complex models, very quantitative models, to make a decision, to make predictions. And we then use very systematic approaches to validate these predictions, preclinically and then clinically, to the point that, very much like the model of the equation of motion that Newton developed, you can now have the equation of cancer, or the equation of diabetes, that you can use to predict whether a particular patient will respond or not respond to a particular drug.
ANDREA CALIFANO: And if they don't respond, how do we sensitize them by using, say, a second drug. And so this is-- the concept of precision medicine is really driving medicine from a more empirical base of trial and error, if you want, driven by certain principle, but relatively-- alludes to something where you have a very, very sophisticated model that is highly predictive.
Segment:6 Do you have an example of the application in precision medicine?.
ANDREA CALIFANO: We think of precision medicine in cancer as targeting a particular protein that is mutated, like an EGF receptor, with a drug called erlotinib or afatinib. And yet the idea is that it sounds great in principle because you say, I have a mutation. I have a drug that targets that mutation. What's more precise than that? The problem is that only about 35% of the patients that have the EGFR mutation, activated mutation, will respond to these drugs, and the vast majority of those will eventually relapse.
ANDREA CALIFANO: Same thing in HER2-positive breast cancer treated with trastuzumab. You got a mutation in a receptor kinase protein. You target it with an antibody. You shut down completely that path. And yet about 30% of the patients don't respond initially and 70% will eventually fail to respond even if they did respond initially. So that's where there's a tremendous need for additional precision.
Segment:7 Do you have anything else to add?.
ANDREA CALIFANO: For a long time, there's been sort of a division between a group of investigators that were called computational biologists and a group that were sort of experimental biologists. And there was a time when I started first in the Computational Biology Group at IBM where we were starved for data. We literally had to go beg for getting data.
ANDREA CALIFANO: We are now at the exact opposite. The experimental biologists are coming to the computational biologists to ask them to process their data because it no longer fits in an Excel table where they can actually see it. But more importantly, I think what's happening-- this is certainly my case. We are starting to develop labs that are fully integrated. So there are labs where we do both the experimental part and the computational part.
ANDREA CALIFANO: And this, I think, is a extraordinary way to simplify the process because you generate the data that you need to create a model. You create the model. That model makes predictions. You can now put those predictions back into the lab and validate them. And from the validation, you can now learn how to improve the model and make it more sophisticated, more accurate.
ANDREA CALIFANO: And so at the end of this process, I hope that we will not think of ourselves as either being experimental or computational biology but rather as just being biologists, or oncologists, or whatever discipline of life science you're interested in.
Segment:8 What are your AACR 2019 highlights?.
ANDREA CALIFANO: It's extraordinarily difficult to talk about AACR highlights because this is such a humongous conference. You know, there's 20,000 people. There's millions of things that are going on. Certainly an area that is becoming of great interest is the use of immuno-oncology approaches. And there is a tremendous spurt of development. I think also there's been a bit of a slowdown in terms of sort of thinking that the mutational space will solve the problem of cancer. And so this has really created an opportunity for these orthogonal approaches based on systems biology and other types of discipline, I think.
ANDREA CALIFANO: AACR likes to call it convergence. I think Phil Sharp come up with that term. And basically sort of bringing together a variety of different disciplines. For instance, before I was talking about information theory, which comes from a branch of mathematics, and biology to come up with models that can actually be predictive of cancer response to certain drugs.
ANDREA CALIFANO: I think this is a very important process that has to go on. Hopefully, the AACR will continue to support that. Another key area, I think, is going to be around doing a better job with models. How do we model cancer so that when we actually go to a clinical study our probability of success increases quite substantially? And this can only be done if more sort of high-fidelity models are used, and also if we really learn to understand what a model is good at.
ANDREA CALIFANO: So for instance, tomorrow in my presentation, I will discuss how we used a cell line that is actually histologically not even from the same type of the tumor that we were studying, which ended up being the most faithful model we could use to actually predict response to drugs in patients that had been validated in actual explants from patients with that particular tumor type. So I think this is, again, an area that is expanding quite dramatically.
ANDREA CALIFANO: [MUSIC PLAYING]
Form Title Download
Transcript
Form Title Bookmark
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Share
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Accessibility and Keyboard Shortcuts
The cadmore media player is a professional video and audio player experience designed to deliver media as well as relevant transcript, segmentation, and metadata. On the screen is a traditional html video player with buttons such as play, pause, forward, captioning, language and play speed selection and more. Adjacent to the video player are generally tabs, one of which shows a scrolling transcript of the video and the other shows any segmentation of the media, sometimes referred to as chapters. At the top of the video is a menu bar containing an Explore button and Tools button. The Explore button allows you to dig deeper into the video by opening up an explore dialog which contains tabs for a visual navigation of the video using thumbnails, an about tab which contains rich metadata about the media, a segments tab which contains further information about the video, and a related tab which contains links to other media related to this media. The tools button opens a menu with options for sharing the media on social media, creating bookmarking links, creating embed codes for your website, generating citations, and more.
Keyboard Shortcuts
Keyboard shortcuts only work while you are in forms mode. By entering forms mode, you will be able to quickly access portions of the media player with single key commands.
Spacebar: Play/Pause Toggle
I: Info Menu,
O: Tools Menu
T: Transcript Tab,
S: Segments Tab
F: Forward 15,
R: Rewind 15
L: Languages,
P: Playspeed
M: Mute Toggle,
V: Volume Bar
N: Video Only Toggle,
B: Full Screen Toggle
C: Captions Toggle
Escape: Closes dialogs and menus / Exits Player
Z: Next Transcript Chapter
X: Previous Transcript Chapter
Q: Next Transcript Paragraph
W: Previous Transcript Paragraph
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Embed Size
-
You Size - Responsive
-
We Size - Responsive
-
Picture Overlay Popup
-
400 x 225
-
512 x 288
-
400 x 700
-
768 x 1024
-
1024 x 512
-
1280 x 608
Form Title Cite
No citation provided.
Info
Segments
Related
Thumbnails
Exit Clip Mode
You have requested to see a portion of the video outside of the clip.
Click OK to switch to the full video or click if you would like to stay inside the clip. You can always go back to the clip with Tools and Clip Mode