Thoughts on Scientific Research Innovation: A Review of Director Pu Muming's Speech at the 2010 Annual Meeting of the Institute of Neurology, Chinese Academy of Sciences

Thoughts on Scientific Research Innovation: A Review of Director Pu Muming's Speech at the 2010 Annual Meeting of the Institute of Neurology, Chinese Academy of Sciences

Introduction: Innovation is a concept that is inseparable from science. Director Pu mainly talked about his views on innovation in detail from the perspective of experimental science (December 28, 2010, according to the audio recording);

First of all, I would like to talk about this annual meeting, which is very unusual. This time, the most foreign units came to participate. In previous years, we only invited one unit at most. This year we invited three units [Institute of Biophysics, Chinese Academy of Sciences, Fudan University Institute of Brain Science, Zhejiang University Medical Center], each group has fewer opportunities to speak, but the content of our topic is richer, and there are more opportunities for communication. I think everyone has some thoughts after listening to the three-day speech , I haven’t decided what to talk about three days ago. After listening to it, I think I need to raise this issue again: innovation, innovative people, and innovative research. We all talk about innovation now, and the whole country talks about it. Innovation, you see innovation being talked about in the newspapers every day. I want to [talk about] my views on innovation. I have said some things many times so far. Wordy, but I am qualified to be wordy now, because I am an old man...

Innovation, what is innovation? Let’s think about it first, what is real innovative research. There are many kinds of innovative research. To do theory, you have a new view on natural phenomena, put forward a new concept, a new Theory, this is the innovation of theoretical scientists. For our biology, most of them are doing experimental science, and there are many kinds of experimental innovations. For example, there is a new phenomenon observed that others have not seen You see, you designed a very ingenious experiment that can review natural phenomena, these natural phenomena can only be seen under your experimental conditions, but others cannot, such as sub-molecular, sub-atomic phenomena, are all Various new technologies in physics, such as phenomena that can only be seen after the accelerator is available. What was not clearly seen in biology before, two-photon can see more clearly, and new microscopy methods can see more clearly. These are innovations , technological innovation, innovation that sees new phenomena. There is another kind of innovation, that is, you did not say that you discovered a new phenomenon, but you clarified the causal relationship between two phenomena that others saw, and proposed a new phenomenon. The relationship between these two phenomena promotes our understanding of the natural world. Others have seen this phenomenon and that phenomenon, but they don’t know the relationship between these two phenomena. If you figure out the relationship, this is also innovation. What else? , Even if you don’t have this, you use the phenomenon seen by others, but you have a new analysis method to explain this phenomenon better, which is also an innovation.

I think now, we often talk about innovation. For many people who have just started doing research, how can I do innovation? Now I can only use the technology I have learned, which are all other people’s technology, and I have seen other people’s phenomena. How can you have the opportunity to innovate even when you are a beginner? I think the simplest example is that you have the data to get it. Generally, you need to analyze the data, whether it is microscopy data or your biochemistry data. What is your method of analysis? You always follow the articles published by others, and you do what they do. This is the most common. In fact, when you are a graduate student and do original research, you should think of On the basis of other people's analysis methods, can you further do better analysis, more quantitative, and even the way of drawing pictures is new. No one else has drawn pictures like this, and you draw this picture is new. I often Some graduate students brought the data and said that I don’t know how to analyze it, ah, this is the best situation! Because you don’t know how to analyze it, it means that others have not done this data, and you have no examples to follow, so you can’t follow what others have done. Diagram method analysis. This is the best time for innovation, the best opportunity; if you want to come up with a method and be able to make an analysis, this is the innovation of the method. There is also drawing, and the most boring thing when you see the paper is that it is all You have not gone further than the method others have done; although it is similar data, it is presented in a new method, which is refreshing, and this is innovation.

Innovation has various levels of innovation. We all do research. For example, we analyze dendrite branch. I see that everyone follows the previous analysis method in this respect, but many details are not considered. For example, look at dendrite length, branch point, you ignore a lot of other parameters, you may ignore a lot of details and didn’t see it; counting the spine number, I didn’t look at the spine intensity before, but now from the perspective of quantification, the spine intensity is actually information. So I’m very happy to see Yu When the students in the Xiang group reported, they thought that spine intensity differences also had information, which is very good. Most of the spine analysis did not do this. Any one more step in analysis is present here, and even the students who are just beginning can try it Yes. Also, you use methods from other fields to your research. In your field, no one else has used this method to conduct experiments. If you bring it in, this is also innovation.

In addition, everyone says that innovation is to find new things, see new phenomena, and acquire new knowledge. This is a kind of innovation. But there is another kind of innovation, which I think is even more important. When I was doing research, I saw the present [at the annual meeting]. The work you do is not much different from other laboratories in the world, laboratories in Europe and America, and you are competitive with them. , and may not be able to compete, because good laboratories, overseas laboratories, they have good resources, good equipment, good students, good environment, you have to compete with them, if If you are just entering this field, then you are at a disadvantage. So I say that there are two types of innovation in research: one kind of innovation is really early, what the previous people have done, and what is the next step, you have to find out The following new knowledge and new methods are innovations; there is another kind of innovation, which is actually not called innovation. I call it "retrospective research". You use the current method to look at old problems, and you don't have to ask With new theories and new phenomena, you can reinvestigate the phenomena that people see now, and use new methods to reinvestigate it. For example, you can solve the controversial issues now, which is also innovation. Although it is not the problem you raised , It’s not the phenomenon you saw, but you used new analysis methods, new experiments. You even brought things from textbooks, those concepts and theories were based on experiments thirty or forty years ago, very old The technology and the superficial data you have seen have drawn conclusions, but they have entered the textbooks (because they are very important), and no one is going to test these problems. You are using modern and new methods to design experiments and repeat old problems, old hypothesis, Check whether they are correct; basically repeat the experiment, but must use a new method. I believe that 50% [experimental results] are confirmed original results, and 50% [original experimental results] are not solid. As you know, on 50% of the things in the average textbook are wrong, need to be revised, but we don’t know which 50%, so you can do this job again, this is a win-win situation. Everyone likes new knowledge, no one Repeat, repeat experiments do not have credit, so people say that the master has already done it 20 or 30 years ago, am I asking for trouble if I do it now? However, it is very likely that you will reopen this question,In fact, this kind of retrospective research is a kind of innovative work, and its results are as important as the early things, because the progress of science is to constantly revise our current knowledge. In the process of revising, you can benefit from a lot of new knowledge and new Phenomena that have accumulated have caused you to have to change your current views. We are generally like this now. When new knowledge begins to accumulate, any discrepancies will be hidden, and as the hypothesis becomes more discrepant to a certain extent, you will The hypothesis must be changed, the current theory must be changed; this is our so-called paradigm shift, Thomas Kuhn’s paradigm shift, and paradigm shift means that we must change after the emergence of new knowledge. But what about the retrospective research method I mentioned? , go directly to attack, go to check the foundation of current opinion, right? Look directly where there is cleft, you can consolidate the foundation either, or overturn it. If you have 50% chance you are to win, the other 50% You are not a loser, because you have confirmed the old phenomenon with a new method, and you can also publish paper, which is also an important paper, because your [result] is an important confirmation. I have been advocating it for at least ten years, that is, go back To the old hypothesis, do experiment, repeat experiment, I can say that the successful things in my own laboratory, several things, are revisit the hypothesis. This is the benefit of revisiting the old hypothesis, you can open a new area, especially for young investigators.When new knowledge begins to accumulate, any discrepancies will be concealed, and if there are more discrepancies with the hypothesis to a certain extent, you have to change the hypothesis, you have to change the current theory; this is what we call paradigm shift, Thomas Kuhn Paradigm shift, paradigm shift means that after new knowledge appears, it must be changed. But what about the retrospective research method I mentioned, go directly to attack and check the foundation of current opinion, right? Just look where there is cleft, You can either consolidate this foundation, or overturn it. If you have a 50% chance you are to win, and the other 50% you are not a loser, because you have confirmed the old phenomenon with a new method, you can also publish paper , is also an important paper, because your [result] is an important confirmation. I have been advocating it for at least ten years, that is, go back to the old hypothesis, do experiment, repeat experiment, and I can say that I have successfully done it in my own laboratory Several things are revisit the hypothesis. This is the benefit of revisit the old hypothesis, you can open a new area, especially for young investigators.When new knowledge begins to accumulate, any discrepancies will be concealed, and if there are more discrepancies with the hypothesis to a certain extent, you have to change the hypothesis, you have to change the current theory; this is what we call paradigm shift, Thomas Kuhn Paradigm shift, paradigm shift means that after new knowledge appears, it must be changed. But what about the retrospective research method I mentioned, go directly to attack and check the foundation of current opinion, right? Just look where there is cleft, You can either consolidate this foundation, or overturn it. If you have a 50% chance you are to win, and the other 50% you are not a loser, because you have confirmed the old phenomenon with a new method, you can also publish paper , is also an important paper, because your [result] is an important confirmation. I have been advocating it for at least ten years, that is, go back to the old hypothesis, do experiment, repeat experiment, and I can say that I have successfully done it in my own laboratory Several things are revisit the hypothesis. This is the benefit of revisit the old hypothesis, you can open a new area, especially for young investigators.You can either consolidate this foundation, or overturn it. If you have a 50% chance you are to win, and the other 50% you are not a loser, because you have confirmed the old phenomenon with a new method, you can also publish paper , is also an important paper, because your [result] is an important confirmation. I have been advocating it for at least ten years, that is, go back to the old hypothesis, do experiment, repeat experiment, and I can say that I have successfully done it in my own laboratory Several things are revisit the hypothesis. This is the benefit of revisit the old hypothesis, you can open a new area, especially for young investigators.You can either consolidate this foundation, or overturn it. If you have a 50% chance you are to win, and the other 50% you are not a loser, because you have confirmed the old phenomenon with a new method, you can also publish paper , is also an important paper, because your [result] is an important confirmation. I have been advocating it for at least ten years, that is, go back to the old hypothesis, do experiment, repeat experiment, and I can say that I have successfully done it in my own laboratory Several things are revisit the hypothesis. This is the benefit of revisit the old hypothesis, you can open a new area, especially for young investigators.

So there are many forms of innovation, not necessarily new. But we say that everyone doing original research, in principle, is new. What we saw at the annual meeting today are experiments that others have not done, but is it very What about innovative research? Most of them are not, they all follow current trends, everyone’s hypothesis is also check, add a little bit of your own, there is no real potential to be really innovative. Why? Because the problem you solve is not a hard question, all It’s the details, the big framework is the most important issue. It’s not that new work is innovative, innovative work must have an element of surprise, an element that raises people’s eyebrows to attract people’s attention, oh why didn’t I think of this phenomenon, seeing you There will be elements of surprise in the results, there are hints, this is innovative research. Why is it like this? Why do people pay attention and are surprised? [Because] I don’t think it’s like this of course, it’s not something that’s not new , The phenomenon in my expectation. That is to say, real innovative work must be connected with importance, innovative sciences must be important, unsolved problem.

What is an important unsolved problem? There are two kinds of unsolved problems. One kind of unsolved problem is that we know that we can solve it one day, a problem that known to be solvable. We said that in the 1990s we needed to sequence human genome, that is an important problem, Everyone knows it. It must be solvable. As long as you do sequence, it will take a long time. At that time, the machine is not so good. But you know: If you have the resources, man power, you can attack the problem. Now we do neuroscience, The important solvable problem is the connectome. In principle, we can map every connection in the brain, what we need to have is hundreds of man power. According to current technology, hundreds of labs work hundreds of years, you will map every branch every connection .The technology is there, how to analyze, how to make sense of it. Now there are many centers in foreign countries to do this, including the center of neural connectome, which needs to map the connections in the cortex. If China wants to do this It’s not impossible, as long as you do it. But I don’t think this is the best way to solve the problem. Genome came out, and it didn’t solve the problem of disease. We still need to know the function of the gene. It is important, but it isn’t t the best problem you should spend your life on it. Because of the human genome,There were no sequencing machines in the 1990s, but now it only takes one month for sequencing. It used to be ten years, so those people spent ten years in the 90s regret, he can use this time to work on more important issues.

The second type of problem has little or no clue how it can be solved. Now there are many problems that have not been solved, why? Because we don’t know how to start. There are some clues that we can start with, language, perception, even more complicated, Consciousness is a very important issue. The most important issue in our neuroscience is to understand the human brain, and the most important thing about the human brain is human specific properties, but we have little clue how to do it. This is the key, these important issues It can be solved, you can find a way, this is where you can use your intellect.

How to choose the question is mentioned below. Peter Madewar said, I think everyone has heard this, I will say it again. Peter Madewar is an immunologist, Nobel prize winner, the wisest man in biology in 20th century to me. He wrote It is very interesting for you to read in our reading room. He said: Science is art of soluble. Here he does not mean how to solve problems is science. What he really means is: It is finding a problem that is " Soluble” at this time by you. What you want to choose is in your hands, and at this time is an important issue of soluble. How to choose? This art cannot be explained clearly in a few words, it needs to be learned in a lifetime. Individuals learn in different ways, get different conclusions, and choose different questions. Everyone has their own style. The difference is that you always have some clues when choosing questions, but how are you willing to take risks for these clues? Do you want to do safe research or risky research? Everyone has a different style and art. Everyone has their own way of doing science. I think this is the most important thing. How to choose, including Now also, you choose a difficult problem, whether it can be solved in your hands. If you feel confident, you can overcome the situation deficiency, you can also take risks, you can choose. This is a big problem, so I want to talk about it In the words of Hu Shi, Hu Shi said to "make bold assumptions and carefully seek proofs", of course he was talking about research. The same goes for doing science, you have to do innovative science, do things that no one else has done, and find new ones. This is like an adventure. In the same way, you have to find a new path. You must take risks during the adventure, otherwise it is not an adventure; if you don’t take risks, you will not get benefits, you will not find treasures, and you will not find new paths. , there are many travelers; if the danger is far away, there will be few people who arrive at it. The world's magnificence, grotesque, and extraordinary views often lie in the danger and distance, and the rareness of people is so rare, so those who are not ambitious can't go there"]. So make a bold assumption,propose your hypothesis, propose your goal. I think of this passage:

“In biology, any mechanism you can imagine, as long as it is useful to the organism and doesn’t violate physical and chemical principles, is likely to be utilized by the organism.” (Poo’s Dictum)

It's your imagination to dream out these hypotheses. This kind of hypothesis cannot be the same as your motivation, violate principles. ( = =b ) As long as it is a reasonable hypothesis, you can pursue it. This is a bold assumption. I have said this sentence many times , I probably heard someone say it, or I read it in a book, so I was looking for this sentence, but I couldn’t find it. Since I can’t find it, I claim it is my own...

You can’t tell where the idea came from, it’s a very important thing, everybody wants to claim intellectual property. But it’s hard, is it what they tell you when you talk to them, or half of what they tell you, you think for yourself Half of it, hard to say. So I think in science, not in technology, intellectual property is nonsense. If everyone wants to think about intellectual property, then the world will be in chaos, and there will be no communication in the scientific community and no real progress. In the past two years If the progress of science in the 10th century is still the same in the 21st century, it will go downhill. The progress of science in the 21st century depends on communication. Until the 1980s and 1990s, molecular biology was developed to commercialize science. [Mr. Pu is indeed a respectable idealist].

How to innovate? I have talked about this in class, and some students may have forgotten it, so I will talk about it again.

(I).Knowing how the facts are obtained is more important than knowing the facts.

How he did it, why he thought of doing this job, he will not say in the paper. Most important findings are unexpected, they are not what he originally wanted to do, [the real purpose] he will not tell you, not in The paper tells you that this is the result of the style of writing the paper and scientific stereotyped essays. But we can know that we often have first-class scientists passing by. I ask the students to ask questions. Next time you can’t think of a question, if you are silent, you will Raise your hand, ask a standard question, just say why do you do that experiment, I tell you, you will get the best answer, he will say Ah, that's interesting! We were doing this thing, and then we find that and finally got it. So [others] innovation process, you have to dig for it. I said why we have lunch with these visitors, just to be able to ask them questions.

There are 5 Ws (What, Who, When, Where, Why), and the last four Ws are history, which is the basis of innovative work. You know how each innovative work comes about, and this has a subtle function .I have told you a lot about this book [The Eighth Day of Creation]. Our first prize today is this book. I have the original version of this book in 1979, which has been preserved for 30 years. It is the first prize this year.

(II).Knowing how to find the facts is more important than knowing how the facts.

There is too much information, and it is not good for you to know the information. Now there is an information overload problem, because too much information will confine your imagination, because every information is like a fact, and it is very certain that I found this, I I found that, so I read a lot of papers, you can’t do anything, you dare not do anything, because your thinking is limited. So do not simply download and read all the paper. The lower the Students in grades should pay more attention to this matter. Before entering a lab, it is most important to read a few review papers. Don’t read all the papers. Read them when you need to read them in the process of doing a job. , don’t download paper there all day long, it’s as harmful to you as smoking. Excess in selected reading is bad for your health. [I think of Pope’s phrase "Too much learning is a dangerous thing."]

(III).Knowing how to present the facts is more important than knowing how the facts.

Our human thinking is very strange. You must train yourself to speak out. The process of speaking is integration. All presentation is a process of logical integration. Linearly presented the causal relationship of your idea, sort out a line, what? What should be said, what to say beforehand, what to say afterwards. General talk is very important. I think the presentation at the annual meeting has made great progress, and everyone is very serious. Science is a social phenomenon. Lock me on an island, Give me endless resources, don't let me talk to others, I can do research, whether I like to be? I won't do research, I would rather go to the sun all day... [Mr. Pu is so cute] I think I No one appreciates what you do, and you have to let people know what you do, so that you will feel that you are doing something useful. So you must have social communication, communicate with others, do it behind closed doors and don’t talk, and immerse yourself in experiments, that’s not enough .And part of fun in science is communication. We know that before molecular biology, a group of 40 people is responsible for moving the molecular biology forward. They have an RNA Tie Club themselves, and everyone who participates in this club has a tie, It is an amino acid, they want to solve coding, how nucleic acid is coded, they are all interested in this problem, they do not publish paper; the function of this Tie Club is unpublished communications, letters, short notes, their fun is here In other places, many important papers of Francis Crick have not been published, and they are all in the communication of RNA Club.

(IV).What’s unknown is more important than what has been known.

You need to know what is an unsolved problem, and what is a soluble problem. All current knowledge is incomplete, problematic, and unreliable. Either correct the questionable problem, or obtain unknown knowledge. The important thing here is to read the review paper; I just said that reading the review paper is very important, because you don’t need to read too many papers. Even if you read the review paper, you have to choose, you need to know which review journal is the best, and write the review Does the person have an overview, is he very experienced in this field? There are many review papers with hundreds of references. The more references, the worse the review paper is, because he doesn’t know how to choose, he puts all reliable and unreliable things Mixed together, he dare not offend people. A really good review is to tell the important things, what is known and what is not known, he summed it up very clearly; there are many things that seem to be known, but he said he does not know , he said we don't know, is debatable. You need to know what is unknown and what is known, don't just choose which paper reference is the most, this will have the opposite effect.

(V).Educating yourself is more important than finding someone to educate you.

We often complain about our environment, there is no Cavendish Lab, where the mentor guides you, and you have a 50% chance of winning the Nobel prize. But the times have progressed, and we don’t need these masters by our side, why? Because first-hand There are a lot of accounts now, and every famous great scientist has written something to tell you how to do innovative science, and it is clearer than what he said next to you, because the things he wrote are all things he has thought about and are very Essence. You can read in our reading room. If you want to be guided by a master, you can go there to study. There are also many autobiographies, how do they make their discovery. And if you don’t want to read books, just want to surf the Internet, then you can go to this website http://webofstories.com, a few students have been on this website, do you like this website? As long as you like science, you can’t get down. There are several hours of interviews. He talks about his history and why he Can do this experiment, what is his experience.

Our time is limited, and there are still ten minutes. I will just talk about a few notes. I just talked about "bold assumptions" and "careful verification". Cautionary notes for innovative research:

• Be rigorous in every step of your work;

• Be receptive to criticism;

• Learn to deal with frustration.

"Bold hypothesis" means that you have to ensure that your adventure will bear fruit, choose a road that no one has taken, and choose the road you should choose. Be rigorous in every step, the conclusion you got in the previous finding is not an experiment If the result is supported, you lose a step, which may lead you to a detour. If you want to get a goal, you must be faithful to your result.

In this process, you have to accept criticism. What is criticism? You have to accept the criticism that the reviewer gives you when the paper goes out. Many people say that the reviewer is an antagonist and wants to suppress you; there may be some people who are not fair, but Most reviewers, especially those from good quality journals, their reviews are fair; they dare not be unfair, because they are not fair, and the next time the journal editor will not let him review, most people will think about their own careers. And Most of the reviewers are useful. I think the biggest achievement of our Institute of Neurology in the past ten years is that the publication has been raised, because the early PIs respect the opinions of the reviewers; if someone criticizes you, you can solve the problem later. Send it back. So be receptive to criticism, even sometimes it is negative. And most people give up.

The third is to learn to deal with frustration. It is not easy to do innovation; if it is easy, others have done it, why do you need to do it? Therefore, to solve important problems, you will definitely encounter many setbacks and spend a lot of time President, the time to get the reward may be very slow, but you have to endure it to achieve the goal. There are many setbacks, such as the frustration of not being able to get the fund, and the frustration of not being able to do experiments. Many people ask me, what is a good one? Scientists, successful scientists, I said that you learn how to learn to deal with frustration, which is the guarantee of success. This is extremely important. I said adversity breeds creativity, which means that without frustration, you can’t make good things. What about a smooth environment? , often fail to motivate you to do your best work. We say that the best works of painters and writers are just at the most difficult time [article hate life], and when he makes a lot of money, his works are not good. Xu Beihong studied painting in Paris When I was hungry, I didn’t even have the money to buy bread. During the Anti-Japanese War, those historical paintings, figure paintings, and five hundred warriors painted in the rear were all in the most difficult time. Why did he draw the best paintings at that time? I think there is a reason, that is, the more difficult you are, somehow, I don’t know why, the more you can stimulate your creativity. You must break through this predicament, and creation will come out. Xu Beihong’s story is still very interesting, he insists on Draw a lot of paintings, because he has to pay, if you want to know the story... [Mr. Pu said a very cold joke below] His former wife must get his paintings, must have hundreds of them painting, he can get a divorce; so in order to get a divorce, he has to paint, and under extreme hardship and frustration, he has painted many good paintings. It’s not over, the paper is generally above the average paper is no problem, it means you need to enter the predicament, you have to find more difficult problems in order to break through this predicament. Some very famous scientists change their fields every ten years, our board member Charles Stevens [UCSD, HHMI, NAS member] is like this, because he thinks that the original field is almost done, there is no challenge, the new problem is the challenge,Challenge is a predicament. If you break through the predicament, you will go to another level [Here we know why not everyone can achieve Mr. Pu’s achievement].

OK, thank you! Happy New Year!