Big data technologies in working with microbiota bacteria. Lecture in Yandex
We often talk about tasks that lie at the junction of a particular classical science and data analysis. In today's report, this ideology is presented firsthand - the scientist reads most of the report, and the programmer talks about specific methods and tools.
Under the cut - the decoding and the main part of the slide.
Dmitry Alekseev, director of R & D at Atlas biomedical holding:
- Actually, I'm a microbe. This is unusual to hear, because you are used to the fact that people tell you something. And I'm a microbe. We live a lot. And in you too.
')
The figure is approximate - 100 trillion each. You can multiply by the number of visitors. So many of us are here today. I have come to tell you something important about us, microbes, because lately you haven’t paid much attention to this, and it will not lead to anything good. It is important that you have two things in your head when I finish: that, firstly, I am good, and secondly, that the data available to you right now from your laptop can tell even more about what we are interesting .
When we, microbes, are many, we are called microbiota. And now there is a great hope in your human medicine that in the near future you will again be able to romp, drink, smoke and eat the wrong way, and we, the microbes, will save you back. And in general, it is now called theranostica - when we diagnose what hurts you, and we will cure it too. We, microbes, generally can do anything. And it is clear that we have been living on this planet much longer than you. From the very moment you appeared, while still being a little worm, we have already populated you. And it is important to understand that all your organs, especially the intestines, about which we are going to talk, - they developed together with my ancestors, although, in truth, I have no ancestors like a microbe. We share in half. A little later you will see this beautiful process.
If you wanted to know whether it is worth doing microbes, and in 2014 you climbed into the database of publications - this is such a logarithmic scale - you would see that they are engaged in microbes. Thousands of articles a year interesting, people write a lot of interesting things about immunity and microbes, and cancer and microbes, and in general about the connection of microbes or microbiota, especially the intestines, with diseases. Hot topic. And this hot topic is the post-genomic era. After people counted, calculated, read the human genome, they turned on the same technologies to microbes. And they found out that there are more of us than expected.
They study us because we have DNA. You will be surprised, but in me the DNA is the same as in you. I just imagine this thought in silence sometimes I think that the DNA in me is the same as in people. It is amazing. Still, we are super-different, but everything is written with zeroes and ones. Although maybe you are very familiar. Why? Because you make all these applications, and useful applications, and useless. And a useful application of zeroes and ones, and useless. The main thing here is to build order correctly. Although I am sure that no one lined up my order, evolution worked for me. Those my brothers who were stronger, the ancestors, they just stayed. Those who somehow incorrectly changed, they disappeared. This is what is called screening. And the totality of our genes is called the meta-gene. This is what you - the person who analyzes the data - is served primarily as raw data. And these genes can be found in any environment. Even in a nuclear reactor or in a space station, there are my microbe friends, and they also have genes, and they act according to the same rules.
If you look into the intestines, you can see the diversity, that is, there are many guys like me. But there are also other species. They are in a different cap, in other glasses, someone in sneakers, someone in shorts. Approximately 300 to 1000 species of such children live in each intestine. And it is surprising: if we estimate this, then we will not really understand which cells are bigger: me, that is, microbes, or your own. But we all fit in these 1.5 kg of dry weight that you have in the intestines. Why? Because we are really less. If you have not guessed, then I am now - just an enlarged copy of the microbe.
Therefore, the number of genes, some functional sections of DNA that can do something, is a couple of orders of magnitude higher, so in fact I can do much more than you, and I am your evolutionary device so that you can be a hooligan. You will never change your genetic code, and when you start eating bad food, first of all I learn to adapt to this. Now I will show you directly from the data how I do it.
Good picture, very old. Then nobody read about genes too hard. Part of the disease in English is written, the incidence is falling. These are all infectious diseases caused by the same guys as me, but they are gopnik-like. And other diseases are related to your immune system. For example, diabetes or asthma. Something in it does not work. And this is happening right nowadays. In humanity, there are fewer infectious diseases, and diseases that are associated with immunity become larger. It is interesting.
What happened, why did the number of infectious diseases decline? Because you invented antibiotics. Antibiotics are such small substances that prevent me from either multiplying or moving, and then I die. And you take these antibiotics in order to get rid of the bad guys, but everyone suffers, because we are organized almost equally, there is no specific medicine for a particular bacterium, a separate microbe. And so the good guys also died - immunity suffers.
I have been studying you, people in the past 10 years: a PhD in biological sciences - this means that I study life. All this time, I think that the human immune system was created in order to calculate which microbes are there, to support them correctly. And basically it happens in the intestines, and, again, by the number of cells. That is, you are perfectly able to control who should be there.
Here is also an interesting story, not yet related to the data. Since childhood, everything depends on how you were born. If this is a caesarean section, then the first microbes get into your intestines from the mother’s skin. If this is the original, primordially human mode of birth, then the vaginal microbiota, and it turns out that our intestines got used to be made right during this time, to grow, based on the fact that at first microbes are transmitted from mother to her vagina. If they are from the skin of the mother, then it is a completely different microbes and starts to happen wrong. This is often expressed in children, for example, in allergies. This is confirmed by already existing recommendations: if there was a caesarean section, then the child is smeared in the mother's juice, and then the right microbes get into the intestines, and then everything happens correctly.
In children, these microbes change a lot: now they are different. They get sick, and immunity gets used. At some point, we think, the window of opportunity closes. You remember all the good and all the bad due to your immune system. This is a supercomputer. She can by one small element, simply by a cap, recognize me as a microbe and say whether she is herself or someone else’s. Therefore, if you drink all sorts of kefir with live bacteria, they will remain with you, but until three or five years it is not written anywhere that they are their own. It is possible to reflect that in those children who are born now, the same code is already embedded, because they have been drinking these synthetic bacteria since childhood. Perhaps this is the place for superhacker attacks on future people, because all of them have the same synthetic bacterium code. You will call the company that does it yourself.
In adulthood, our combination, our ratio is about as you eat. So it is arranged. Why do you need a lot of different? Because we are experts in different things. Experts to digest different polymers in your food. Basically, you used to eat not so much animal food, a lot of vegetable food, and we, your microbes, have adapted to digest this vegetable food. Each of us can do one thing. Imagine that you eat little plant food, or get it straight from the factory. She is milled, we have almost nothing left. And the experts leave. Just the everyday life of your intestines is always monochrome food from the same cardboard packaging.
When you grow old, we actually think that we are preparing you for death - so that you can digest again into simpler components, and we create a new life out of you on this planet. In principle, we always did that.
An interesting way to understand how it works and how it is built into the human body. There is such a theory that through immune cells from the mother’s intestines, microbes are transmitted with milk to her baby. There, in his intestines, these immune cells open. Thus, at some point, the mother colonizes the intestines of the child with beneficial microbes. In other words, while this evolutionary human body was created, we found a way to get to your children as quickly and safely as possible.
This picture is really interesting. Someday in reality you needed to get rid of a lot of bad germs. You did two things. I already called antibiotics. The second is about hygiene: you made everyone wash their hands. The last option is Mayakovsky, who was afraid of everything at all, and washed everything with his wash. This is a supercultural phenomenon. We must understand that in no living nature of such a phenomenon and meme there, as if everything should be washed. You created it, created some super-tales, and they may be different in different nations. In my childhood, in your childhood, in someone else's childhood in the 1980s, this character Moydodyr definitely was, and the boy with dirty hands and dirty feet was just super-mild. Now, if you put up a “theory of hygiene” on Wikipedia, you will immediately find the theory of a European doctor that the number of allergies in children is due to the fact that they are placed in super-sterile conditions. Due to this, the immune system is not trained, that is, it lives in a sterile box. And then, when faced with something really hostile or useful, she reacts to all this with inflammation. And inflammation is an allergy.
Cool story about food, about how it can be interestingly arranged. For example, you eat meat, and one of my colleagues - also a microbe - can get a substance out of this meat, which will then create cholesterol plaques. If you do not live such a colleague, then you have no problems with meat. If he settled with you, it would be good to think about this, how to regulate it somehow. Because I understand that one of your goals is to stay in this body for as long as possible, while maintaining its efficiency and joy to life. It is necessary that the body was not just a shell that travels for hundreds of years somewhere, but that it was a living shell. Therefore, it is interesting to pay attention. And here all data that is on the Internet already begins.
You began to actively study me during this story about epidemiological obesity in the United States of America. From 1990 to 2008, in some states, obesity as a diagnosis rose from 10% to 30%. In fact, this is not very similar to those Hollywood films that you watched. On the streets really go very fat people. 30% of them are diagnosed. At the same time, over the past 30 years, the nutritional structure of the Americans has not changed. And Jeffrey Gordon, a researcher, found that people who are more slender, slim, and people full with a diagnosis of "obesity" are different microbes. Moreover, when he transplanted these microbes from people with obesity into mice, it turned out that mice gain weight faster. That is, this variant of microbes, which lives in people with obesity, produces more calories from the same amount of food. And then you realized that the composition of microbes has an influence on how you will feel.
The same interesting story happens in the food market when people get meat. If subtherapeutic doses of antibiotics are given to animals, then they slowly gain weight. If you are in a capitalist country, the more you get the weight you sell in dollars from the same amount of food and the same amount of time, the more profitable you will be. Naturally, at some point, all animal producers began to use this property. When these animals get on the table and further inside the person, antibiotics as substances practically do not disappear anywhere, and the person becomes the same consumer of subtherapeutic doses of these antibiotics. And the result is exactly the same - people just gaining weight. Subtherapeutic doses - this is important here, because doses are so small that neither the Federal Service for Health Supervision or the American Consumer Control Authority simply does not detect them. There are some limits of sensitivity.
Your other story is really a living project, you can look at it. He was called Happy Meal. You can guess why. On day 137, food left on the air looks about the same as it looked in the store. Why does she look like that? Because neither the microbes nor the mold do not eat this food. Imagine that such food gets inside you, and there are the same microbes. They also do not really want to eat it. They say, then McDonald's issued a refutation that there is no special chemistry there, just a lot of salt. But if you think about it, we do not care that there in large numbers does not allow microbes to grow normally.
Another picture is already more complicated. And it is already cool - about the data, which in different economic groups show how the per capita income is growing, and what happens with the consumption of meat, empty calories - this is mainly alcohol and sweets - and in general with the number of calories. Here is the economic group A - the richest countries from 1961 to 2009. Income has increased, meat consumption has increased. In group B it started lower, but the same thing happens there. And in fact, all that we see in all economies is if per capita income grows, then people begin to eat more meat, more empty and dietary calories. In addition to India, they do not eat meat there. That is, it is clear from data that the more we earn, the more we eat. And we eat those very empty calories that are not super useful to our microbes. And the way of life, moreover, that we earn more, becomes not super mobile.
We once looked at these microbes in Russia and found that in large cities - Novosibirsk, St. Petersburg - microbes are the same as the Europeans and the Americans. The same combinations. And in the villages of other microbes, some unique formulations. And then we understood that this is a direct effect of globalization, that when a person comes from a village to a city, he falls under the filter of antibiotics that are in food, and simple preservatives, which are in all food. Ao all over the world the same people buy products in all stores. And it filters and removes some microbes, and only those that can survive such genocide remain. Then we all become the same, we cannot distinguish a European from an American and from a Russian who lives in a big city.
Another little story. Everything I said about antibiotics applies to her too. It is important to know this because it is a global topic that can also be traced in the data. In general, you are releasing more and more antibiotics, and the dose from which the first bacterium died, when the first antibiotic was taken, must now be a hundredfold to kill the bacterium. We always adapt due to the fact that we can share once an hour, and it is easy for us to die. We adapt much faster. It is likely that there will be such a bad guy among us, which will not affect any of your antibiotics. It will be transmitted quickly and will be fatal. This is a problem of antibiotic safety. Antibiotics are circulating throughout the medium.
The guys and I made such a picture about the different levels of resistance to different antibiotics in different countries. It can be found online or in this article. Interestingly, the more industrialized a country, such as China, the more resistance. And for the same Chinese, we also see resistance to chemicals. That is, microbes begin to process some unpleasant chemicals that enter the body. They are already adapting. But it is not very clear whether the Chinese are doing it better or not.
A good project in which we participate from the Russian side is the microbiota of the subways of all cities. Now there are almost 100 cities. It all started with New York. We collected DNA from all stations and looked at the microbes. And, for example, about the data - which, by the way, are also open - we recently solved the following problem: can we determine from the composition of microbes which station it is? Because from different stations, different microbes travel to the center, and there everything is mixed on the transplant.
In general, we perceive this topic about microbes as complexity science. It seems to me that the guru of complexity science is in Santa Fe. There is such a Complexity University, Institute of Complexity Science, I once had an internship there. And one of the ways to study complexity ... We call complexity a state from simple objects, between which there are some connections. And the youth system, we have already created from two bacteria, three substances that they exchange, some kind of food that comes in, and intestines. We even pretended at some point. Red and blue bacteria - intestines, and all this communicates with each other. This is called Agent Based Modeling. You can run this story, and then see if we get something similar to what is happening in real life. , , , . . , , .
, , - , , , . .
— , . , Data Science Day. . 200 . k- — k. , -, . . , , - , . - , Bioinformatics. . , , , , , .
. , . . , , , - ---, .
, : , , , . , , . ? . .
, , , .
, - «»:
— , . : Data Science , . , , .
— , — text mining .
, . . , . — - , . , , . PICRUSt. , . . , . , , , . data scientists — — , exploratory- , . , , . , , — .
heatmap. , , B9. , , . , . heatmap , , , . . , , data scientist : ? - . , , 40, .
. , , . , , WGS (whole genome sequencing — ) 9 10 . , 9 10 . . , Big Data, , , HBase — . , Big Data- , . «» . , .
:
— . MetaMut. , , . , , , . 20 , , . , , - , , , . . , , . , , , … , data scientist, , . — , — , - . .
:
— text mining . 16 . . , , , , . . : , . natural language processing —, spaCy, . , , , positive negative. , — , . . , . transfer learning pipeline: - , simple classifier.
. - , - — , Faecalibacterium prausnitzii — - . , . , . , , , , - . PCA.
:
— . . , , . , , , , , , . - -. -. - , , .
, , , . , . , , , , . - — , . , , , . : , , , . — — , .
Thank you very much. I hope you will become kinder to us, microbes, and programmers, of course.
Under the cut - the decoding and the main part of the slide.
Dmitry Alekseev, director of R & D at Atlas biomedical holding:
- Actually, I'm a microbe. This is unusual to hear, because you are used to the fact that people tell you something. And I'm a microbe. We live a lot. And in you too.
')
The figure is approximate - 100 trillion each. You can multiply by the number of visitors. So many of us are here today. I have come to tell you something important about us, microbes, because lately you haven’t paid much attention to this, and it will not lead to anything good. It is important that you have two things in your head when I finish: that, firstly, I am good, and secondly, that the data available to you right now from your laptop can tell even more about what we are interesting .
When we, microbes, are many, we are called microbiota. And now there is a great hope in your human medicine that in the near future you will again be able to romp, drink, smoke and eat the wrong way, and we, the microbes, will save you back. And in general, it is now called theranostica - when we diagnose what hurts you, and we will cure it too. We, microbes, generally can do anything. And it is clear that we have been living on this planet much longer than you. From the very moment you appeared, while still being a little worm, we have already populated you. And it is important to understand that all your organs, especially the intestines, about which we are going to talk, - they developed together with my ancestors, although, in truth, I have no ancestors like a microbe. We share in half. A little later you will see this beautiful process.
If you wanted to know whether it is worth doing microbes, and in 2014 you climbed into the database of publications - this is such a logarithmic scale - you would see that they are engaged in microbes. Thousands of articles a year interesting, people write a lot of interesting things about immunity and microbes, and cancer and microbes, and in general about the connection of microbes or microbiota, especially the intestines, with diseases. Hot topic. And this hot topic is the post-genomic era. After people counted, calculated, read the human genome, they turned on the same technologies to microbes. And they found out that there are more of us than expected.
They study us because we have DNA. You will be surprised, but in me the DNA is the same as in you. I just imagine this thought in silence sometimes I think that the DNA in me is the same as in people. It is amazing. Still, we are super-different, but everything is written with zeroes and ones. Although maybe you are very familiar. Why? Because you make all these applications, and useful applications, and useless. And a useful application of zeroes and ones, and useless. The main thing here is to build order correctly. Although I am sure that no one lined up my order, evolution worked for me. Those my brothers who were stronger, the ancestors, they just stayed. Those who somehow incorrectly changed, they disappeared. This is what is called screening. And the totality of our genes is called the meta-gene. This is what you - the person who analyzes the data - is served primarily as raw data. And these genes can be found in any environment. Even in a nuclear reactor or in a space station, there are my microbe friends, and they also have genes, and they act according to the same rules.
If you look into the intestines, you can see the diversity, that is, there are many guys like me. But there are also other species. They are in a different cap, in other glasses, someone in sneakers, someone in shorts. Approximately 300 to 1000 species of such children live in each intestine. And it is surprising: if we estimate this, then we will not really understand which cells are bigger: me, that is, microbes, or your own. But we all fit in these 1.5 kg of dry weight that you have in the intestines. Why? Because we are really less. If you have not guessed, then I am now - just an enlarged copy of the microbe.
Therefore, the number of genes, some functional sections of DNA that can do something, is a couple of orders of magnitude higher, so in fact I can do much more than you, and I am your evolutionary device so that you can be a hooligan. You will never change your genetic code, and when you start eating bad food, first of all I learn to adapt to this. Now I will show you directly from the data how I do it.
Good picture, very old. Then nobody read about genes too hard. Part of the disease in English is written, the incidence is falling. These are all infectious diseases caused by the same guys as me, but they are gopnik-like. And other diseases are related to your immune system. For example, diabetes or asthma. Something in it does not work. And this is happening right nowadays. In humanity, there are fewer infectious diseases, and diseases that are associated with immunity become larger. It is interesting.
What happened, why did the number of infectious diseases decline? Because you invented antibiotics. Antibiotics are such small substances that prevent me from either multiplying or moving, and then I die. And you take these antibiotics in order to get rid of the bad guys, but everyone suffers, because we are organized almost equally, there is no specific medicine for a particular bacterium, a separate microbe. And so the good guys also died - immunity suffers.
I have been studying you, people in the past 10 years: a PhD in biological sciences - this means that I study life. All this time, I think that the human immune system was created in order to calculate which microbes are there, to support them correctly. And basically it happens in the intestines, and, again, by the number of cells. That is, you are perfectly able to control who should be there.
Here is also an interesting story, not yet related to the data. Since childhood, everything depends on how you were born. If this is a caesarean section, then the first microbes get into your intestines from the mother’s skin. If this is the original, primordially human mode of birth, then the vaginal microbiota, and it turns out that our intestines got used to be made right during this time, to grow, based on the fact that at first microbes are transmitted from mother to her vagina. If they are from the skin of the mother, then it is a completely different microbes and starts to happen wrong. This is often expressed in children, for example, in allergies. This is confirmed by already existing recommendations: if there was a caesarean section, then the child is smeared in the mother's juice, and then the right microbes get into the intestines, and then everything happens correctly.
In children, these microbes change a lot: now they are different. They get sick, and immunity gets used. At some point, we think, the window of opportunity closes. You remember all the good and all the bad due to your immune system. This is a supercomputer. She can by one small element, simply by a cap, recognize me as a microbe and say whether she is herself or someone else’s. Therefore, if you drink all sorts of kefir with live bacteria, they will remain with you, but until three or five years it is not written anywhere that they are their own. It is possible to reflect that in those children who are born now, the same code is already embedded, because they have been drinking these synthetic bacteria since childhood. Perhaps this is the place for superhacker attacks on future people, because all of them have the same synthetic bacterium code. You will call the company that does it yourself.
In adulthood, our combination, our ratio is about as you eat. So it is arranged. Why do you need a lot of different? Because we are experts in different things. Experts to digest different polymers in your food. Basically, you used to eat not so much animal food, a lot of vegetable food, and we, your microbes, have adapted to digest this vegetable food. Each of us can do one thing. Imagine that you eat little plant food, or get it straight from the factory. She is milled, we have almost nothing left. And the experts leave. Just the everyday life of your intestines is always monochrome food from the same cardboard packaging.
When you grow old, we actually think that we are preparing you for death - so that you can digest again into simpler components, and we create a new life out of you on this planet. In principle, we always did that.
An interesting way to understand how it works and how it is built into the human body. There is such a theory that through immune cells from the mother’s intestines, microbes are transmitted with milk to her baby. There, in his intestines, these immune cells open. Thus, at some point, the mother colonizes the intestines of the child with beneficial microbes. In other words, while this evolutionary human body was created, we found a way to get to your children as quickly and safely as possible.
This picture is really interesting. Someday in reality you needed to get rid of a lot of bad germs. You did two things. I already called antibiotics. The second is about hygiene: you made everyone wash their hands. The last option is Mayakovsky, who was afraid of everything at all, and washed everything with his wash. This is a supercultural phenomenon. We must understand that in no living nature of such a phenomenon and meme there, as if everything should be washed. You created it, created some super-tales, and they may be different in different nations. In my childhood, in your childhood, in someone else's childhood in the 1980s, this character Moydodyr definitely was, and the boy with dirty hands and dirty feet was just super-mild. Now, if you put up a “theory of hygiene” on Wikipedia, you will immediately find the theory of a European doctor that the number of allergies in children is due to the fact that they are placed in super-sterile conditions. Due to this, the immune system is not trained, that is, it lives in a sterile box. And then, when faced with something really hostile or useful, she reacts to all this with inflammation. And inflammation is an allergy.
Cool story about food, about how it can be interestingly arranged. For example, you eat meat, and one of my colleagues - also a microbe - can get a substance out of this meat, which will then create cholesterol plaques. If you do not live such a colleague, then you have no problems with meat. If he settled with you, it would be good to think about this, how to regulate it somehow. Because I understand that one of your goals is to stay in this body for as long as possible, while maintaining its efficiency and joy to life. It is necessary that the body was not just a shell that travels for hundreds of years somewhere, but that it was a living shell. Therefore, it is interesting to pay attention. And here all data that is on the Internet already begins.
You began to actively study me during this story about epidemiological obesity in the United States of America. From 1990 to 2008, in some states, obesity as a diagnosis rose from 10% to 30%. In fact, this is not very similar to those Hollywood films that you watched. On the streets really go very fat people. 30% of them are diagnosed. At the same time, over the past 30 years, the nutritional structure of the Americans has not changed. And Jeffrey Gordon, a researcher, found that people who are more slender, slim, and people full with a diagnosis of "obesity" are different microbes. Moreover, when he transplanted these microbes from people with obesity into mice, it turned out that mice gain weight faster. That is, this variant of microbes, which lives in people with obesity, produces more calories from the same amount of food. And then you realized that the composition of microbes has an influence on how you will feel.
The same interesting story happens in the food market when people get meat. If subtherapeutic doses of antibiotics are given to animals, then they slowly gain weight. If you are in a capitalist country, the more you get the weight you sell in dollars from the same amount of food and the same amount of time, the more profitable you will be. Naturally, at some point, all animal producers began to use this property. When these animals get on the table and further inside the person, antibiotics as substances practically do not disappear anywhere, and the person becomes the same consumer of subtherapeutic doses of these antibiotics. And the result is exactly the same - people just gaining weight. Subtherapeutic doses - this is important here, because doses are so small that neither the Federal Service for Health Supervision or the American Consumer Control Authority simply does not detect them. There are some limits of sensitivity.
Your other story is really a living project, you can look at it. He was called Happy Meal. You can guess why. On day 137, food left on the air looks about the same as it looked in the store. Why does she look like that? Because neither the microbes nor the mold do not eat this food. Imagine that such food gets inside you, and there are the same microbes. They also do not really want to eat it. They say, then McDonald's issued a refutation that there is no special chemistry there, just a lot of salt. But if you think about it, we do not care that there in large numbers does not allow microbes to grow normally.
Another picture is already more complicated. And it is already cool - about the data, which in different economic groups show how the per capita income is growing, and what happens with the consumption of meat, empty calories - this is mainly alcohol and sweets - and in general with the number of calories. Here is the economic group A - the richest countries from 1961 to 2009. Income has increased, meat consumption has increased. In group B it started lower, but the same thing happens there. And in fact, all that we see in all economies is if per capita income grows, then people begin to eat more meat, more empty and dietary calories. In addition to India, they do not eat meat there. That is, it is clear from data that the more we earn, the more we eat. And we eat those very empty calories that are not super useful to our microbes. And the way of life, moreover, that we earn more, becomes not super mobile.
We once looked at these microbes in Russia and found that in large cities - Novosibirsk, St. Petersburg - microbes are the same as the Europeans and the Americans. The same combinations. And in the villages of other microbes, some unique formulations. And then we understood that this is a direct effect of globalization, that when a person comes from a village to a city, he falls under the filter of antibiotics that are in food, and simple preservatives, which are in all food. Ao all over the world the same people buy products in all stores. And it filters and removes some microbes, and only those that can survive such genocide remain. Then we all become the same, we cannot distinguish a European from an American and from a Russian who lives in a big city.
Another little story. Everything I said about antibiotics applies to her too. It is important to know this because it is a global topic that can also be traced in the data. In general, you are releasing more and more antibiotics, and the dose from which the first bacterium died, when the first antibiotic was taken, must now be a hundredfold to kill the bacterium. We always adapt due to the fact that we can share once an hour, and it is easy for us to die. We adapt much faster. It is likely that there will be such a bad guy among us, which will not affect any of your antibiotics. It will be transmitted quickly and will be fatal. This is a problem of antibiotic safety. Antibiotics are circulating throughout the medium.
The guys and I made such a picture about the different levels of resistance to different antibiotics in different countries. It can be found online or in this article. Interestingly, the more industrialized a country, such as China, the more resistance. And for the same Chinese, we also see resistance to chemicals. That is, microbes begin to process some unpleasant chemicals that enter the body. They are already adapting. But it is not very clear whether the Chinese are doing it better or not.
A good project in which we participate from the Russian side is the microbiota of the subways of all cities. Now there are almost 100 cities. It all started with New York. We collected DNA from all stations and looked at the microbes. And, for example, about the data - which, by the way, are also open - we recently solved the following problem: can we determine from the composition of microbes which station it is? Because from different stations, different microbes travel to the center, and there everything is mixed on the transplant.
In general, we perceive this topic about microbes as complexity science. It seems to me that the guru of complexity science is in Santa Fe. There is such a Complexity University, Institute of Complexity Science, I once had an internship there. And one of the ways to study complexity ... We call complexity a state from simple objects, between which there are some connections. And the youth system, we have already created from two bacteria, three substances that they exchange, some kind of food that comes in, and intestines. We even pretended at some point. Red and blue bacteria - intestines, and all this communicates with each other. This is called Agent Based Modeling. You can run this story, and then see if we get something similar to what is happening in real life. , , , . . , , .
, , - , , , . .
— , . , Data Science Day. . 200 . k- — k. , -, . . , , - , . - , Bioinformatics. . , , , , , .
. , . . , , , - ---, .
, : , , , . , , . ? . .
, , , .
, - «»:
— , . : Data Science , . , , .
— , — text mining .
, . . , . — - , . , , . PICRUSt. , . . , . , , , . data scientists — — , exploratory- , . , , . , , — .
heatmap. , , B9. , , . , . heatmap , , , . . , , data scientist : ? - . , , 40, .
. , , . , , WGS (whole genome sequencing — ) 9 10 . , 9 10 . . , Big Data, , , HBase — . , Big Data- , . «» . , .
:
— . MetaMut. , , . , , , . 20 , , . , , - , , , . . , , . , , , … , data scientist, , . — , — , - . .
:
— text mining . 16 . . , , , , . . : , . natural language processing —, spaCy, . , , , positive negative. , — , . . , . transfer learning pipeline: - , simple classifier.
. - , - — , Faecalibacterium prausnitzii — - . , . , . , , , , - . PCA.
:
— . . , , . , , , , , , . - -. -. - , , .
, , , . , . , , , , . - — , . , , , . : , , , . — — , .
Thank you very much. I hope you will become kinder to us, microbes, and programmers, of course.
Source: https://habr.com/ru/post/331158/
All Articles