Informatics submarine in the steppes of biology
Bioinformatics is rapidly gaining popularity and is turning from a geek shelter into a well-known well-established discipline. I think most readers of Geektimes can confidently say that a rabbit is not only valuable fur and 3-4 kilograms of dietary meat, but also 44 chromosomes, a variety of different proteins, transcription and translation mechanisms, and anything else. Also, I can hardly surprise anyone if I say that all this can be studied and analyzed not only standing in a white robe at a microscope in a sterile laboratory, but also lying on the couch with a laptop, drinking something Scottish with ice. However, beyond this knowledge usually do not go. I decided to try to correct this unfortunate misunderstanding and take a short tour of how bioinformatics looks from the inside from a practical point of view , based on my experience.
In this article I will collect the questions that I myself asked three years ago, when I was still a student at the Faculty of Mathematics, and I will try to answer them.
The task of bioinformatics, informally speaking, is to find the logic in biological data. These data are obtained in the course of experiments, and if the data for a biologist can look like a glowing fish or a beautiful multi-colored spot in a photo, then for bioinformatics data are presented as:
')
And many other possible representations of real biological phenomena with the help of mathematical objects.
Sure. But bioinformatics do not need to run to the laboratory on weekends (cell cultures, for example, do not know about the weekend and tend to die without proper care). Yes, and research in biology often lasts for years (depending on the properties of model organisms), while in bioinformatics progress depends mainly on the ability to solve algorithmic problems and write “smart” code. Well, the possibility of remote work from anywhere in the world is also a definite plus in favor of bioinformatics.
It very much depends on the specific research center and research group. To understand biology at the minimum level you need - no one will chew on a scientific project to the level of a school math problem. You yourself will need to model the situation based on your understanding of biology. However, a truly deep understanding is not expected, so the fact that you only remember about the pistils and stamens will not be an obstacle if you decide to take up this particular science. The necessary basics of biology is easy to learn already in the process of working on a bioinformatics project.
What is really useful and necessary for the future of bioinformatics “from informatics” is knowledge of biotechnology, that is, how your data was obtained, which problems could arise during the experiment. In my opinion, it is enough to run a gallop on some course of molecular biology, but to spend time and seriously comprehend the principles of operation of modern instruments used for experiments.
I would advise the future bioinformatics “from biology” to skip first proofs and descriptions of methods and algorithms and study them as “black boxes”, that is, in a purely applied aspect: “A at the input - B at the output”, otherwise there is a risk “ "drown" in theoretical calculations for several years. However, having missed the theory and having learned something in practice, it will be easy for you to go back and look at it with different eyes.
Unfortunately not. Bioinformatics in its current state is a lot of fairly voluminous sections, as in any other science. If we compare, for example, with physics, it is rather obvious that a specialist in theoretical mechanics will most likely experience certain difficulties in understanding recent articles on quantum physics, and moreover, he most likely will not have time to read these articles.
And there are a lot of sections in bioinformatics and for every taste:
and much more (forgive me for bioinformatics, whose field I forgot to mention).
The last three points are often referred to systems biology, but these sciences, as they say, “at the junction”, can be jumped back and forth with minimal effort.
To answer this question, rank the following characteristics according to their significance for you (give rank 6 to the most important characteristic, 1 - the least important), and then summarize with the indicated sign.
+ I always wanted to be a scientist and feel that I make a definite contribution to the future of mankind.
+ I am interested in life sciences, I would like to have the opportunity to learn something new about biology every day, but my university studies were not related to biology - or - I am a biologist, but I was bored with monotonic technical manipulations with pipettes and I want more understand exactly what data I received and be able to work with them.
+ Bioinformatics is interesting to me as a subsection of computer science, it seems to me that there are many tasks that need to be thought of.
- I want to get a big salary immediately after graduation from the university.
- I would like to constantly walk in a white coat, like a real scientist.
- I like to think about tasks and read interesting articles on biology, but I do not like programming.
If you get a result less than 0, you definitely should not go into bioinformatics. You feel the pain of how lax and not universal this test, but you understand his idea and even like something? Add yourself +3 points to the result.
“If you really want to, you can fly into space,” but if you are 2 meters tall and weighing 150 kg, you are unlikely to be taken into the cosmonaut corps. And what about bioinformatics?
Career is laid out with higher education. Undergraduate can be any, but still not humanitarian. Economics, physics, chemistry, mathematics, not to mention computer science and biology.
The most favorable choice of master's degree is either a master's degree in bioinformatics, or a “supplement” to your undergraduate degree, so that you have something biological, and something computational after these two steps. True, to enter the magistracy with a completely different profile is not an easy task.
As for the possibility of obtaining the first stage of higher education (bachelor / specialist) immediately with a specialization in bioinformatics - I have an ambiguous attitude towards this.
Bioinformatics should be a conscious choice, and making such a choice after school looks rather difficult, but if you are sure that this is your vocation, then why not. I’m more impressed with the approach of “getting a general education and then choosing a specialization” rather than immediately starting to work in a narrow direction. I am not sure that it is easy to retrain to a specialist in another profile after 4-6 years of training, but there are examples of successful informatization.
Quite a few online courses have already been created for the capturing acquaintance with bioinformatics (Russian-language Stepic.org, English-language Coursera, edX, etc.). Among the online courses there are very useful (I would recommend a course from UCSD on algorithms in bioinformatics and a course on evolution from Duke University), sign up, go through if it becomes boring or difficult - you calmly give up this case without spending any other time or your nerves. After all, for full-time study it is decent to go motivated - figuratively speaking, in a three-piece suit, with a bouquet in your hands and a carnation in the buttonhole - so that bioinformatics will immediately understand that this relationship is serious for you.
Additional education is a wonderful thing, which has only some advantages - classes on weekends or evenings (does not interfere with basic studies or work), an enthusiastic team and often even the lack of tuition fees. But - the selection for such programs is quite tough, the courses are voluminous and the pace is fast. That is why, if you still just want to understand whether to be engaged in bioinformatics further, it’s better to do so before - watch online courses, talk to people from the profession, read something popular science (from what people wrote, which I know personally, - article on Habrahabr , article on Geektimes , review “I would go to bioinformatics - let them teach me” on the Biomolecule).
As far as I know, there are two additional programs in Russia - at the Institute of Bioinformatics (IB) in St. Petersburg and in the capital - the Moscow School of Bioinformatics (MSHB) . In my opinion, they are approximately equal to the master's degree in the level of knowledge gained in their specialty, but only “a rare bird reaches the middle of the Dnieper” - many students fall off, having visited a dozen classes - oh, hard work is to collect the hippo genome.
I myself graduated from the Institute of Bioinformatics as part of the master program at the Academic University (SPbAU) , so I’ll tell you more about information security (I don’t know anything about MSHB after they parted with Yandex). The program lasts a year, classes on Saturdays. Seminars and lectures I liked almost everything, but the most remarkable part of the training is scientific projects. Supervisors there are from the leading scientific centers of Russia and the treacherous abroad. In theory, projects should be educational in the first place, but most often this is the real science. It was a good time: sleepless nights full of Arabic tales “1000 and 1 script” (in fact, at first the tales were Hindu), fierce defenses of the projects, and a feeling of belonging to the very leading edge from which scientific articles often come meet on geektimes. Yes, by the way, there is a buffet. And the set goes there now . At the same time, both the dignity and lack of information security is the absence of fundamental disciplines - only bioinformatics, and nothing else.
If you want more subjects and fundamental training, then the holders of technical bachelor / specialist diplomas can, like me, immediately go to the 2-year magistracy in algorithmic bioinformatics . The admission process is standard: online applications until mid-summer, then an interview. Call for applications for 2016/18 is already open . But biologists go there completely pointless.
For the sake of completeness, I had to use my network of agents. On the eve of the publication, one of the intelligence officers finally broke the radio silence mode and handed over a radiogram about MSB to headquarters. The main points in the decrypted message about the learning process in the MSHB were: a) the possibility of obtaining an official diploma from HSE; b) the presence of fundamental disciplines like matan (in my opinion, this is a mockery, but matan is useful because it leads the mind in order); c) research projects are carried out under the guidance of leading bioinformatists of Moscow; d) because of the abundance of homework, students have to flock together and think collectively about problems; e) students, however, are squealing with delight and asking for more bioinformatics. Recruitment in the MSHB will begin in May.
Another type of additional education. From what I know: there is a School of Molecular and Theoretical Biology for schoolchildren (more on biology, but for the future bioinformatics the benefits are unquestionable), for students and “beginner” graduate students - the Summer School of the Institute of Bioinformatics (LSHB) , from foreign - Research Summer School in Statistical Omics (RSSSO) . If it’s very brief about those schools where I was - LSH is ideal for a short intensive introduction to bioinformatics, RSSSO is for those who already understand what computational biology is and wants to “pump” their statistical base. At LShB / RSSSO you can / should take part in interesting research projects, during which you can feel for a short time a real scientist. In addition, a great way to have fun in the summer in a great company. LSH is held alternately in Moscow and St. Petersburg, RSSSO - in Croatia, the city of Split. ShMTB will be in Barcelona.
Then begins, in fact, a career - after the graduate you can get a job as a bioinformator (yes, yes, I hear outraged voices, after the bachelor’s degree, after school, and after kindergarten, but let's agree that graduation is the best many parameters point of reference). This can be done both in Russia (a database of vacancies is collected on the website blastim.ru) and abroad. The second option is to go get a PhD or Ph.D. Find a graduate school (in almost any country - even in Russia, at least in Costa Rica) is quite simple, provided that you are a good specialist. Diploma grades play a role, but not a decisive one. Where is better - abroad or at home? Suspend this question. Perhaps, by the time you are ripe for admission to graduate school, you will decide for yourself. All the same, in the process of studying in graduate school, you will most likely be trained in one country or several times in another country for several months.
After Ph.D. There are already 3 options:
The first is to understand that life is decay, quit science altogether and go to the Yamalo-Nenets Autonomous District to breed deer. We will not dwell on this option, since it is no longer associated with the topic of the article (but I would advise to beware of wolves and not anger deer, their horns look rather dangerous).
The second option is to continue an academic career, and the third is to go into the industry (many companies are now looking for specialists of the appropriate profile). Academic career involves obtaining several internships, which are called, abbreviated, postdok. Salaries for postdocs are several times higher than for graduate students, but, as a rule, they are lower than the salaries of those specialists who go into the industry. Find a job in the industry after receiving a Ph.D. and (optional) several postdocs are much easier. Then you can get a permanent position of a researcher or try to create your own laboratory and lead it. This is a complicated matter and, to be honest, what happens “behind the postdoc” I do not know anything.
I will continue to answer your questions asked in the comments to this article. Also, if there is interest in that, I can tell you about what I do (study of the links between genetic and epigenetic variability and diseases) in a separate article.
About the author: Specialist, Faculty of Mechanics and Mathematics of Moscow State University, 2013; Master of Science (Bioinformatics), MIIT SPbAU, 2015; currently a graduate student at CRG, Barcelona, ​​a group of “Genomic and Epigenomic Variation in Disease”.
I hope this text was informative for you.
PS Before sending this article, read by colleagues, and stated that it was written too pessimistically. I dare to assure readers that, asking these questions a few years ago, I received much more gloomy answers.
In this article I will collect the questions that I myself asked three years ago, when I was still a student at the Faculty of Mathematics, and I will try to answer them.
Why do we need bioinformatics?
The task of bioinformatics, informally speaking, is to find the logic in biological data. These data are obtained in the course of experiments, and if the data for a biologist can look like a glowing fish or a beautiful multi-colored spot in a photo, then for bioinformatics data are presented as:
')
- strings (sequences of characters describing DNA / RNA / proteins);
- three-dimensional and two-dimensional coordinates (microscopy data);
- arrays of real numbers (for example, each number can be an experimentally measured mass of a protein or part of it);
- vectors of nonnegative integers (for example, the depth of coverage with discrete objects, the so-called reads );
- matrices of zeros and ones (for example, can different types of bacteria get along with each other);
And many other possible representations of real biological phenomena with the help of mathematical objects.
Do biologists have more interesting data?
Sure. But bioinformatics do not need to run to the laboratory on weekends (cell cultures, for example, do not know about the weekend and tend to die without proper care). Yes, and research in biology often lasts for years (depending on the properties of model organisms), while in bioinformatics progress depends mainly on the ability to solve algorithmic problems and write “smart” code. Well, the possibility of remote work from anywhere in the world is also a definite plus in favor of bioinformatics.
How many in bioinformatics bio, and how much - informatics?
It very much depends on the specific research center and research group. To understand biology at the minimum level you need - no one will chew on a scientific project to the level of a school math problem. You yourself will need to model the situation based on your understanding of biology. However, a truly deep understanding is not expected, so the fact that you only remember about the pistils and stamens will not be an obstacle if you decide to take up this particular science. The necessary basics of biology is easy to learn already in the process of working on a bioinformatics project.
What is really useful and necessary for the future of bioinformatics “from informatics” is knowledge of biotechnology, that is, how your data was obtained, which problems could arise during the experiment. In my opinion, it is enough to run a gallop on some course of molecular biology, but to spend time and seriously comprehend the principles of operation of modern instruments used for experiments.
I would advise the future bioinformatics “from biology” to skip first proofs and descriptions of methods and algorithms and study them as “black boxes”, that is, in a purely applied aspect: “A at the input - B at the output”, otherwise there is a risk “ "drown" in theoretical calculations for several years. However, having missed the theory and having learned something in practice, it will be easy for you to go back and look at it with different eyes.
But if I become a bioinformatist, then I will know bioinformatics?
Unfortunately not. Bioinformatics in its current state is a lot of fairly voluminous sections, as in any other science. If we compare, for example, with physics, it is rather obvious that a specialist in theoretical mechanics will most likely experience certain difficulties in understanding recent articles on quantum physics, and moreover, he most likely will not have time to read these articles.
And there are a lot of sections in bioinformatics and for every taste:
- Evolution (not only in the form of “first were Pithecanthropes”, but also lesser-known issues, such as the evolution occurring in a cancer tumor)
- Search for genetic variants leading to diseases
- Design and selection of drugs that bind to certain types of “dangerous for the body” proteins
- The study of the functions of genes, their annotation
- Structural bioinformatics (manipulations with 2D and 3D structures, such as, for example, proteins or RNA)
- Genome assembly
- Mapping of how this whole mix of proteins / RNA / DNA / fat / smart thoughts / gym / Kremlin diet and other things reacts to each other (just like in this video , but more interesting and more difficult)
- Simulation of complex systems (such as the development of the organism from the embryo)
- Neurobiology (or rather, the analysis of data obtained by neuroscientists);
and much more (forgive me for bioinformatics, whose field I forgot to mention).
The last three points are often referred to systems biology, but these sciences, as they say, “at the junction”, can be jumped back and forth with minimal effort.
Does it make sense to choose bioinformatics as your profession?
To answer this question, rank the following characteristics according to their significance for you (give rank 6 to the most important characteristic, 1 - the least important), and then summarize with the indicated sign.
+ I always wanted to be a scientist and feel that I make a definite contribution to the future of mankind.
+ I am interested in life sciences, I would like to have the opportunity to learn something new about biology every day, but my university studies were not related to biology - or - I am a biologist, but I was bored with monotonic technical manipulations with pipettes and I want more understand exactly what data I received and be able to work with them.
+ Bioinformatics is interesting to me as a subsection of computer science, it seems to me that there are many tasks that need to be thought of.
- I want to get a big salary immediately after graduation from the university.
- I would like to constantly walk in a white coat, like a real scientist.
- I like to think about tasks and read interesting articles on biology, but I do not like programming.
If you get a result less than 0, you definitely should not go into bioinformatics. You feel the pain of how lax and not universal this test, but you understand his idea and even like something? Add yourself +3 points to the result.
What does a career ladder look like for bioinformatics?
“If you really want to, you can fly into space,” but if you are 2 meters tall and weighing 150 kg, you are unlikely to be taken into the cosmonaut corps. And what about bioinformatics?
Basic education
Career is laid out with higher education. Undergraduate can be any, but still not humanitarian. Economics, physics, chemistry, mathematics, not to mention computer science and biology.
The most favorable choice of master's degree is either a master's degree in bioinformatics, or a “supplement” to your undergraduate degree, so that you have something biological, and something computational after these two steps. True, to enter the magistracy with a completely different profile is not an easy task.
As for the possibility of obtaining the first stage of higher education (bachelor / specialist) immediately with a specialization in bioinformatics - I have an ambiguous attitude towards this.
Bioinformatics should be a conscious choice, and making such a choice after school looks rather difficult, but if you are sure that this is your vocation, then why not. I’m more impressed with the approach of “getting a general education and then choosing a specialization” rather than immediately starting to work in a narrow direction. I am not sure that it is easy to retrain to a specialist in another profile after 4-6 years of training, but there are examples of successful informatization.
Additional education
Quite a few online courses have already been created for the capturing acquaintance with bioinformatics (Russian-language Stepic.org, English-language Coursera, edX, etc.). Among the online courses there are very useful (I would recommend a course from UCSD on algorithms in bioinformatics and a course on evolution from Duke University), sign up, go through if it becomes boring or difficult - you calmly give up this case without spending any other time or your nerves. After all, for full-time study it is decent to go motivated - figuratively speaking, in a three-piece suit, with a bouquet in your hands and a carnation in the buttonhole - so that bioinformatics will immediately understand that this relationship is serious for you.
Additional education is a wonderful thing, which has only some advantages - classes on weekends or evenings (does not interfere with basic studies or work), an enthusiastic team and often even the lack of tuition fees. But - the selection for such programs is quite tough, the courses are voluminous and the pace is fast. That is why, if you still just want to understand whether to be engaged in bioinformatics further, it’s better to do so before - watch online courses, talk to people from the profession, read something popular science (from what people wrote, which I know personally, - article on Habrahabr , article on Geektimes , review “I would go to bioinformatics - let them teach me” on the Biomolecule).
As far as I know, there are two additional programs in Russia - at the Institute of Bioinformatics (IB) in St. Petersburg and in the capital - the Moscow School of Bioinformatics (MSHB) . In my opinion, they are approximately equal to the master's degree in the level of knowledge gained in their specialty, but only “a rare bird reaches the middle of the Dnieper” - many students fall off, having visited a dozen classes - oh, hard work is to collect the hippo genome.
I myself graduated from the Institute of Bioinformatics as part of the master program at the Academic University (SPbAU) , so I’ll tell you more about information security (I don’t know anything about MSHB after they parted with Yandex). The program lasts a year, classes on Saturdays. Seminars and lectures I liked almost everything, but the most remarkable part of the training is scientific projects. Supervisors there are from the leading scientific centers of Russia and the treacherous abroad. In theory, projects should be educational in the first place, but most often this is the real science. It was a good time: sleepless nights full of Arabic tales “1000 and 1 script” (in fact, at first the tales were Hindu), fierce defenses of the projects, and a feeling of belonging to the very leading edge from which scientific articles often come meet on geektimes. Yes, by the way, there is a buffet. And the set goes there now . At the same time, both the dignity and lack of information security is the absence of fundamental disciplines - only bioinformatics, and nothing else.
If you want more subjects and fundamental training, then the holders of technical bachelor / specialist diplomas can, like me, immediately go to the 2-year magistracy in algorithmic bioinformatics . The admission process is standard: online applications until mid-summer, then an interview. Call for applications for 2016/18 is already open . But biologists go there completely pointless.
For the sake of completeness, I had to use my network of agents. On the eve of the publication, one of the intelligence officers finally broke the radio silence mode and handed over a radiogram about MSB to headquarters. The main points in the decrypted message about the learning process in the MSHB were: a) the possibility of obtaining an official diploma from HSE; b) the presence of fundamental disciplines like matan (in my opinion, this is a mockery, but matan is useful because it leads the mind in order); c) research projects are carried out under the guidance of leading bioinformatists of Moscow; d) because of the abundance of homework, students have to flock together and think collectively about problems; e) students, however, are squealing with delight and asking for more bioinformatics. Recruitment in the MSHB will begin in May.
Summer schools
Another type of additional education. From what I know: there is a School of Molecular and Theoretical Biology for schoolchildren (more on biology, but for the future bioinformatics the benefits are unquestionable), for students and “beginner” graduate students - the Summer School of the Institute of Bioinformatics (LSHB) , from foreign - Research Summer School in Statistical Omics (RSSSO) . If it’s very brief about those schools where I was - LSH is ideal for a short intensive introduction to bioinformatics, RSSSO is for those who already understand what computational biology is and wants to “pump” their statistical base. At LShB / RSSSO you can / should take part in interesting research projects, during which you can feel for a short time a real scientist. In addition, a great way to have fun in the summer in a great company. LSH is held alternately in Moscow and St. Petersburg, RSSSO - in Croatia, the city of Split. ShMTB will be in Barcelona.
Bioinformatics career
Then begins, in fact, a career - after the graduate you can get a job as a bioinformator (yes, yes, I hear outraged voices, after the bachelor’s degree, after school, and after kindergarten, but let's agree that graduation is the best many parameters point of reference). This can be done both in Russia (a database of vacancies is collected on the website blastim.ru) and abroad. The second option is to go get a PhD or Ph.D. Find a graduate school (in almost any country - even in Russia, at least in Costa Rica) is quite simple, provided that you are a good specialist. Diploma grades play a role, but not a decisive one. Where is better - abroad or at home? Suspend this question. Perhaps, by the time you are ripe for admission to graduate school, you will decide for yourself. All the same, in the process of studying in graduate school, you will most likely be trained in one country or several times in another country for several months.
After Ph.D. There are already 3 options:
The first is to understand that life is decay, quit science altogether and go to the Yamalo-Nenets Autonomous District to breed deer. We will not dwell on this option, since it is no longer associated with the topic of the article (but I would advise to beware of wolves and not anger deer, their horns look rather dangerous).
The second option is to continue an academic career, and the third is to go into the industry (many companies are now looking for specialists of the appropriate profile). Academic career involves obtaining several internships, which are called, abbreviated, postdok. Salaries for postdocs are several times higher than for graduate students, but, as a rule, they are lower than the salaries of those specialists who go into the industry. Find a job in the industry after receiving a Ph.D. and (optional) several postdocs are much easier. Then you can get a permanent position of a researcher or try to create your own laboratory and lead it. This is a complicated matter and, to be honest, what happens “behind the postdoc” I do not know anything.
Instead of conclusion
I will continue to answer your questions asked in the comments to this article. Also, if there is interest in that, I can tell you about what I do (study of the links between genetic and epigenetic variability and diseases) in a separate article.
About the author: Specialist, Faculty of Mechanics and Mathematics of Moscow State University, 2013; Master of Science (Bioinformatics), MIIT SPbAU, 2015; currently a graduate student at CRG, Barcelona, ​​a group of “Genomic and Epigenomic Variation in Disease”.
I hope this text was informative for you.
PS Before sending this article, read by colleagues, and stated that it was written too pessimistically. I dare to assure readers that, asking these questions a few years ago, I received much more gloomy answers.
Source: https://habr.com/ru/post/390563/
All Articles