Encouraging Innovation

Top

00:00:06.21 Hello, my name is Tony Hyman.
00:00:08.28 I'm the director at the Max Planck Institute
00:00:11.04 of Molecular Cell Biology and Genetics
00:00:12.25 in Dresden, Germany.
00:00:14.04 Today, I'd like to talk to you about encouraging innovation
00:00:16.17 in the biomedical sciences.
00:00:18.08 I'd like to focus on, how can we best
00:00:20.06 support young scientists
00:00:22.08 during their first independent career.
00:00:24.18 So as a way of introduction,
00:00:27.07 let's step back and think of the following.
00:00:28.21 Why do societies actually fund basic research?
00:00:31.04 It's because of the understanding that
00:00:33.11 basic research is often the motor for innovation.
00:00:36.20 There are many examples of this
00:00:38.19 both now and through history.
00:00:40.03 But as a perfect example for biomedical
00:00:41.20 sciences, is the huge cluster of biotech
00:00:44.02 and pharma companies in the bay area.
00:00:46.20 Which has been greatly supported by
00:00:48.27 the basic research from UCSF,
00:00:50.23 Berkeley, and Stanford.
00:00:52.10 I think we all understand that if those
00:00:54.11 great universities didn't exist,
00:00:55.25 then it's very unlikely that there would be a cluster of biotech
00:00:58.27 or pharma companies in the bay area.
00:01:01.13 Another aspect of basic research
00:01:04.04 is the understanding that innovation
00:01:05.26 tends to correlate with youth.
00:01:07.26 And here's one example,
00:01:09.19 which is that biomedical scientists tend
00:01:12.09 to cite innovative research topics
00:01:14.06 early on in their career.
00:01:15.09 And as they get older, it tends to cool down.
00:01:17.01 Another metric we can use, is to look at
00:01:20.07 the Nobel prize in medicine.
00:01:21.19 The prize winning work
00:01:23.05 for the Nobel prize has been carried out
00:01:25.00 on average, at the age of 40.
00:01:27.09 And that hasn't changed very much
00:01:28.19 over the last 150 years.
00:01:30.13 And yet, if you look at the American system,
00:01:32.23 the age of the first RO1 grant
00:01:36.05 is now on average, 42.
00:01:38.05 So let's look at those statistics carefully.
00:01:41.11 This shows the percentage of young
00:01:43.19 NIH investigators funded from 1980
00:01:46.24 to today, and you can see that in
00:01:48.21 1980, roughly 20% of investigators
00:01:51.19 were under 36, which gave about 2600 grants.
00:01:55.03 That systematic decline until now
00:01:57.28 we're at about 600, which is about 2-3%
00:02:01.04 of investigators. And so we systematically
00:02:02.21 taking money from young people
00:02:06.07 and giving it to old people.
00:02:08.28 Now the NIH has used a number of
00:02:12.02 different metrics to try and solve this problem.
00:02:14.08 But so far, they haven't really
00:02:17.04 got very far, as you can see,
00:02:18.15 in changing these overall demographics.
00:02:20.15 So why do young people have so much problem?
00:02:22.21 And I think it's because young scientists
00:02:24.19 are always having to look over their shoulders.
00:02:26.17 For instance, a key problem is misguided evaluation metrics.
00:02:30.26 Such as journal impact factor.
00:02:32.17 And the citation index.
00:02:34.02 These are not good ways to encourage innovation
00:02:36.28 because by definition, most of the innovative work
00:02:40.00 is not highly cited right at the beginning,
00:02:41.24 that tends to come later.
00:02:42.23 A key problem is they're competing with senior laboratories
00:02:46.08 for the same resources.
00:02:47.11 Senior laboratories, because they're longer
00:02:49.24 and more established, tend to have
00:02:52.01 networks in the community,
00:02:53.28 they have networks in the meeting organizers,
00:02:55.26 to the journal editors,
00:02:58.16 and everyone else who's involved in organizing science.
00:03:01.00 It's often quite hard for a young person coming in
00:03:02.19 to compete with those resources.
00:03:04.09 A key problem is that a young scientist needs
00:03:07.00 preliminary data to get jobs and grants.
00:03:08.29 And so they have to carry on with the work
00:03:11.08 they did as a postdoc.
00:03:13.02 And that by definition is not as innovative
00:03:16.22 as if they could go and start something completely new
00:03:19.07 in their own laboratory.
00:03:20.19 Once they have their laboratory,
00:03:21.29 they're in a constant state of review,
00:03:24.03 which of course leads to short-term research.
00:03:26.03 Because if you want long-term research
00:03:28.11 with lofty goals, you have to
00:03:29.27 be reviewed over longer time periods.
00:03:32.17 And why is this a particular problem?
00:03:35.19 Is that it means that career and discovery
00:03:38.12 now are becoming conflated.
00:03:40.16 And scientists, especially young scientists today,
00:03:42.29 are beginning to see their career as more important
00:03:45.08 than discovery, because that's what they're rewarded for.
00:03:47.16 So what I'd like to talk about now
00:03:50.24 is some ideas that have come out of Europe
00:03:52.27 on to better support young scientists
00:03:54.26 at the beginning of their career.
00:03:56.15 And this is my own experience,
00:03:58.19 which is that I did my PhD
00:03:59.29 in the Laboratory of Molecular Biology
00:04:01.14 with it's many Nobel prizes.
00:04:02.28 And I, of course, got to see innovation in action there.
00:04:05.06 And I had a very good postdoc in
00:04:07.16 the University of California, San Francisco
00:04:09.02 where I became sort of intimately acquainted
00:04:11.17 with the American system and funding,
00:04:13.02 before moving to EMBL in Heidelberg.
00:04:15.29 Now, EMBL, was a laboratory
00:04:18.25 that was not only doing great research
00:04:20.28 but was also a laboratory for how to fund research.
00:04:23.21 How to fund young people.
00:04:25.13 And I then took the ideas from the EMBL
00:04:28.16 to help set up my own institute
00:04:30.01 in Dresden, which was the Max Planck
00:04:31.14 Institute for Molecular Biology and Genetics.
00:04:34.10 Now, why was EMBL such a success?
00:04:38.02 It was because it had this new model for funding young scientists.
00:04:42.16 When it was set up in the 70s,
00:04:44.09 this was very radical at the time,
00:04:46.04 the idea was to give young people
00:04:48.01 nine years of funding
00:04:49.19 and come back and see what they discovered.
00:04:51.23 The EMBL in Heidelberg has about 50 group leaders
00:04:54.07 and a few senior scientists just to make sure
00:04:56.17 there is a little bit of continuity,
00:04:57.17 but fundamentally, it's a fishing pond
00:05:00.02 for young scientists to come in
00:05:01.06 and do what they want.
00:05:02.21 And the average starting age for a group leader,
00:05:05.08 since the beginning of the laboratory, is 34.
00:05:07.28 So, these people have been given
00:05:09.23 independent funding for nine years
00:05:12.05 to get on and do whatever they want.
00:05:15.04 Now, by any manner or means,
00:05:18.19 this has been successful.
00:05:20.11 There are group leaders from the EMBL all over
00:05:22.19 the world in successful leadership
00:05:25.01 positions. And importantly, it's in the top 10
00:05:27.15 research institutes worldwide
00:05:29.05 based on publication quality.
00:05:30.24 Even though almost everyone leaves
00:05:33.13 after nine years.
00:05:35.00 Why did it work so well?
00:05:38.08 Well, one of the key things was,
00:05:41.06 it gave scientists the freedom to innovate.
00:05:43.11 Of course, at a very young age,
00:05:45.11 they were given a decade of funding
00:05:47.15 to do what they wanted.
00:05:48.19 And that's sort of the obvious aspect of the EMBL.
00:05:51.16 But there is a second and more important issue
00:05:53.26 here, which is that it separated the funding of
00:05:56.16 junior and senior scientists.
00:05:58.28 Now, if you remember, I talked at the beginning
00:06:01.17 of one of the key problems of junior scientists
00:06:03.20 face in the standard granting system
00:06:05.05 is that when a grant comes in, they're compared
00:06:07.10 to more senior colleagues.
00:06:08.29 But the EMBL system avoided that
00:06:10.26 by basically only funding young group leaders
00:06:13.21 who had to compete with each other.
00:06:15.11 So, a key step forward in European science
00:06:19.00 was taken when the EU decided to set up
00:06:22.02 a new funding body called the ERC.
00:06:25.05 Now, the EU has funded science for many decades,
00:06:29.01 basically under a contract mechanism
00:06:31.22 where you had deliverables and milestones
00:06:34.02 you had to reach and there'd be a successful network
00:06:35.24 of scientists to get out of Europe.
00:06:37.22 But of course, as it was standing, it was too goal driven.
00:06:40.06 And so a few people got together
00:06:43.14 in the early 2000s and started
00:06:45.28 to think about how to fund a new science agency.
00:06:48.23 In which excellence was the only evaluation criteria.
00:06:52.06 And this led to the founding of the European Research Council,
00:06:55.06 officially started in 2007.
00:06:57.13 With a budget of about $13 billion.
00:06:59.23 Now the important thing about ERC
00:07:01.27 was that they set up this new funding mechanism
00:07:03.18 which had three sets of grants.
00:07:06.08 Starting, consolidator, and advanced grants.
00:07:09.08 Now, this separated the funding
00:07:11.29 into different career stages.
00:07:13.14 So the Starting Grants
00:07:15.06 were about 2-7 years after the PhD
00:07:17.09 and then the Consolidator Grants
00:07:20.04 which are 7-12 years
00:07:22.06 and then the Advanced Grants.
00:07:23.19 And the important thing about this, of course,
00:07:25.09 is that each stage in the career
00:07:26.26 can be evaluated differently.
00:07:28.17 You can evaluate starting PIs according to
00:07:31.10 the metrics you'd like.
00:07:32.23 Which is, do you have a new innovative idea
00:07:35.08 that you would like to try out in your own lab?
00:07:36.22 And the Consolidator Grant
00:07:38.25 you can think about,
00:07:40.27 have they run a successful lab? And if so,
00:07:42.15 can we then support this longer to get the ideas going?
00:07:46.00 And the Advanced Grants,
00:07:47.23 are of course senior scientists who have a record,
00:07:51.03 a track record of achievement,
00:07:52.27 but would like to shift direction.
00:07:54.15 And so each of these individual cases,
00:07:56.07 you can provide different evaluation criteria.
00:07:58.21 And so this has been very successful.
00:08:02.14 If you look at the average starting age
00:08:04.28 of ERC starting grants,
00:08:07.02 it's about 35.
00:08:09.18 And over half of all ERC funding
00:08:12.01 is devoted to investigators within 12 years of their PhD.
00:08:16.21 The question is, could you apply the ERC
00:08:22.21 system to other systems in the world?
00:08:24.29 And how do you do that?
00:08:26.07 Well, in fact, many countries
00:08:28.03 have a kernel of these systems to support young people.
00:08:31.14 As an example, in the United States,
00:08:34.14 there's NIH Director's New Innovator Award.
00:08:36.13 Which is almost exactly the same as an ERC
00:08:39.01 Starting Grant.
00:08:39.29 You fund risky, innovative research
00:08:42.04 within 10 years of the PhD.
00:08:43.15 It's roughly the same grant size and duration
00:08:46.25 as an ERC Starting Grant.
00:08:48.02 The age is roughly similar.
00:08:49.18 But, only 40-60 are awarded every year.
00:08:53.24 What you could imagine, is adapting the U.S. system
00:08:58.24 to fund young people
00:09:00.13 by expanding the Innovator Award to fund
00:09:02.15 a thousand young investigators per year.
00:09:05.01 Now, how did we come up with the number 1000?
00:09:08.09 Isn't that rather a lot?
00:09:09.18 Well, you could just do it by looking at statistics.
00:09:12.07 Let's go back and look at our graph
00:09:13.29 about funding of age in different investigators
00:09:18.08 at different stages.
00:09:19.25 So, there were 2600 investigators
00:09:23.07 in 1980 under the age of 36
00:09:25.22 and now there are 600.
00:09:27.20 But, if you take this number 2600,
00:09:31.07 and look at it, actually the number of
00:09:35.29 investigators has doubled.
00:09:37.11 So we'd actually expect 5000 people in the current system
00:09:40.19 to be under 36, but only 600 are.
00:09:44.03 So if you think you'd like to get up to that 5000
00:09:47.11 in a 5 year grant cycle, you'd like to be funding
00:09:49.18 roughly 1000 every year.
00:09:52.05 And so, the proposal that you could imagine
00:09:55.21 would be to expand the New Innovator Award
00:09:57.01 system to fund 1000 young investigators
00:09:58.28 per year, not 50.
00:10:00.29 And this would give a steady-state of about 5000.
00:10:02.16 The important thing is, that if you do the
00:10:04.24 numbers roughly, this is about the same amount of money
00:10:08.15 the ERC is putting into funding young people.
00:10:11.28 And the EU has roughly the same amount of people as the United States.
00:10:15.17 And you get roughly into parity with the
00:10:18.02 system, but the numbers aren't so important to say
00:10:21.01 there needs to be a system where most
00:10:22.25 young people coming out of their postdocs
00:10:24.19 are funded on the system where they themselves
00:10:27.16 are evaluated according to the metrics you'd like to use
00:10:30.18 for starting scientists with their own labs.
00:10:33.08 This is not a new idea, actually.
00:10:36.26 Bruce Alberts proposed it in 2009
00:10:39.00 in this Science editorial.
00:10:40.22 Well, at that stage, you could say
00:10:43.23 no one had done the experiments.
00:10:45.01 And now the ERC's come along
00:10:47.03 I think that's really done experiments
00:10:48.26 to test what Bruce proposed at the time,
00:10:51.10 which is if we increase the number of grants
00:10:54.14 for young people, we'll do much better
00:10:56.19 at getting the funds into young people
00:10:58.19 so that they can get started with their own research.
00:11:01.00 So, just to sort of conclude what I've said,
00:11:06.22 the root of innovation is well known.
00:11:09.14 You have to take smart young people
00:11:11.07 with a record of innovation
00:11:12.22 during their training and give them resources
00:11:15.04 for a considerable period of time
00:11:16.19 and let them discover something.
00:11:18.01 Tell them they'll be evaluated on innovation
00:11:20.17 and of course, make decisions earlier.
00:11:23.19 There will be some mistakes.
00:11:25.05 But there will be more innovation.
00:11:27.16 Just to finish, I'd like to come back
00:11:30.16 and concentrate on the point that of course
00:11:32.23 innovation also requires freedom to fail.
00:11:35.26 Science is about trial and error.
00:11:38.10 One tries some experiments and they work
00:11:40.20 or they don't work.
00:11:41.09 And that can often go on for a few years.
00:11:43.19 And Fred Sanger was a senior scientist
00:11:47.18 in my laboratory as a graduate student
00:11:50.05 at the LMB, and he won two Nobel prizes.
00:11:52.11 One for DNA sequencing, and one for protein sequencing.
00:11:54.16 And he was still there when I arrived as a graduate student
00:11:57.11 as a very respected figure,
00:11:58.07 and he liked to talk to the students.
00:11:59.13 And one time he said to me,
00:12:00.21 he said, "Tony, you know what?
00:12:02.00 95% of all my experiments fail
00:12:03.17 and that's probably an underestimate."
00:12:05.15 But of course, he still went on to win his Nobel prizes.
00:12:08.15 And a second important point to say
00:12:11.07 is, let's take the bureaucracy
00:12:13.02 away from young people.
00:12:14.24 Let's not overload them with bureaucracy
00:12:16.06 because that also stifles innovation.
00:12:18.19 And the last quote I'd like to finish on
00:12:20.07 is from Max Perutz,
00:12:21.04 who is actually the founding director
00:12:23.03 of the Laboratory of Molecular Biology.
00:12:24.29 And he said once that,
00:12:28.15 because of success we had with all the Nobel prizes,
00:12:30.12 earnest young men, in those days,
00:12:32.21 would come around with clipboards
00:12:34.17 to try and find out what our strategy was.
00:12:37.06 And I would say to them, it was quite simple,
00:12:39.07 we find smart young people,
00:12:41.01 we give them resources,
00:12:42.20 and we get out of their way.
00:12:44.20 So, I think in general,
00:12:47.10 moving resources back to funding young people
00:12:51.04 would be a simple and effective strategy
00:12:52.23 to bring innovation back into the hands
00:12:56.02 of young people.
00:12:57.06 And importantly, provide young people
00:12:59.02 with the hope that they can go on and have
00:13:01.08 a career in science and really make
00:13:02.25 fundamental discoveries which is of course
00:13:05.05 the reason why we're all scientists.
00:13:07.19 Thank you very much for listening.

This Talk

Speaker: Anthony Hyman

Audience:

Educators of H. School / Intro Undergrad
Student
Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: July 2016

Talk Overview

Tony Hyman presents the case that most innovative research is done by young scientists early in their careers. In the USA, however, NIH funding for scientists under the age of 36 has dropped significantly for the last several decades. Hyman argues that expanding grants that are available only to young scientists and provide 5-10 years of funding would greatly encourage innovation in biomedical research.

Speaker Bio

Anthony Hyman

Tony Hyman received his BSc in Zoology from University College, London and his PhD from the Laboratory of Molecular Biology in Cambridge. He then moved to the University of California, San Francisco to pursue postdoctoral research. Hyman returned to Europe in 1993 when he joined the EMBL in Heidelberg as a young faculty member. After… Continue Reading