With respect to research capability, our aim is to identify and support the most able and dedicated researchers at key career stages and give them access to the best possible equipment and infrastructure.
This is easy to state, but not so easy to achieve. Think first about supporting the best researchers at key career stages.
It is not Wellcome’s job to establish a scientific career structure that runs from cradle to grave in the UK and low- and middle-income countries. But we can’t achieve our objectives if the brightest people see scientific research as a poor career choice and decide to enter a different profession.
Wellcome is not alone in thinking about how best to support research careers. Quite recently, a group called Rescuing Biomedical Research was formed in the USA in response to an article by Bruce Alberts, Marc Kirschner, Shirley Tilghman and Harold Varmus.
The group identified six problems, which can be summarised as:
There are too many researchers and not enough money. This leads to a competitive environment that favours unadventurous research.
Scientists become independent investigators too late in their careers. The average age in the USA (and the UK) is 37 to 38.
Scientists think that funders (especially the NIH) value translational research over discovery research.
Too much time is spent on administration, including grant writing and manuscript revision.
Too much emphasis is placed on the importance of publishing in high-impact journals.
Institutions fail to make firm commitments to faculty members.
Wellcome has identified – and is already addressing – many of these points. To take numbers 2 to 5:
the average age of a new Henry Dale fellow is 34 (still too old, in my opinion, but better than 37)
In addition, we’ve signed the San Francisco Declaration on Research Assessment; we emphasise at every panel meeting that it’s the science that counts, not where it’s published; and our new Senior Research Fellow regulations require institutions to fund renewals in partnership and to support Fellows’ careers.
A distorted career structure
But the real sticking point is too many researchers and not enough money. It’s bad enough that this leads to a hypercompetitive environment and unadventurous research, but an equally serious consequence is a distorted career structure. Most PhD students are able to get a postdoctoral position but many fewer postdocs are able to become independent group leaders.
And those postdocs who fail to become group leaders are often so senior – frequently in their 40s – that it’s difficult for them to identify another career path.
Certainly, except for some large research institutes, there are very few permanent staff scientist positions available, and very few opportunities to run science technology platforms. This uncertainly bedevils a career in biomedical science and probably deters some of the brightest people from applying in the first place.
We need a balanced system
What to do? Even if money were available, the problem cannot be solved by throwing more cash at the existing system, not least because any future funding cuts would exacerbate the problem. Rather, the objective has to be to bring the system into a sustainable equilibrium.
This suggestion was (inevitably) controversial. I agree with the concerns expressed by others, including Kimble and colleagues, that a reduction in scientific expertise in the workforce in general, at a time when scientific expertise is so important, would be counterproductive.
A reduction in PhD numbers might also decrease diversity – bad in itself and also bad for science. As Patrick Vallance has said, tough problems do not get solved by monolithic thinking.
The alternative, highlighted by Kimble and colleagues, is to reduce numbers of postdocs – to make the step from PhD to postdoc harder but to increase the chance that a postdoc will be successful in the search for a group leader position.
This suggestion too will be controversial, not least because it may lead to a reduction in lab sizes. With this in mind, part of the argument will undoubtedly invoke the also-controversial suggestion that beyond a certain point, increasing levels of funding to an individual group leader yields diminishing returns in terms of scientific output.
The alliance of science
I haven’t addressed here how we identify the most able and dedicated researchers, or team science. I’ll write about these in future blogs.
But I want to mention the question of giving researchers access to the best possible equipment and infrastructure. At the moment Wellcome achieves this through three routes:
Biomedical Resource and Technology Development Grants
Multi-user Equipment Grants.
Are there other, or better, routes than these?
Throughout my career I’ve been struck by the importance to researchers of well-resourced science technology platforms, like those at the Crick. I’m keen to make platforms of this sort more readily accessible and more sustainable, and one way of doing this may be to share facilities more widely.
Along these lines, I was struck recently by the shared cryo-electron microscopy facility at the University of Bristol that brings together the Universities of Bath, Bristol, Cardiff and Exeter – the so-called GW4 Alliance.
I hope there will be opportunities to set up other shared regional resources, and simultaneously provide a career path for PhDs who do not want to undertake postdoctoral work?
The appliance of science
Over the (now distant) New Year break I took the opportunity to look at work published towards the end of 2017.
I was struck by a beautiful piece of theoretical work from my former home, the Gurdon Institute, together with the Cavendish Laboratory in Physics Department at Cambridge. Led by Edouard Hannezo and Ben Simons, the work concludes that branching morphogenesis, which occurs in the development of many organs, occurs through a self-organising process that obeys simple generic rules: epithelial cells follow no predetermined genetic programme.
Applications of CRISPR-Cas9 could feature in each of my blogs. In a recent experiment, the technology was used to alleviate symptoms in a mouse model of Amyotrophic lateral sclerosis (ALS), a neurodegenerative condition also known as Lou Gehrig’s disease. In the model, mice express a mutant form of superoxide dismutase 1 (SOD1), a mutation which is responsible for about 20% of ALS cases in humans. Introducing a CRISPR-Cas9 construct into mouse spinal cords reduced mutant SOD1 protein levels – treated mice had 50% more motor neurons and lived 25% longer than controls.
Elsewhere, in an effort to tackle antibiotic resistance, the CryPTIC project is collecting and analysing more than 100,000 tuberculosis (TB) infection samples from across Africa, Asia, Europe and the Americas. It’s funded by a Wellcome Collaborative Award and led by Derrick Crook. Last year this team launched a citizen science project to examine over 40 million images of Mycobacterium tuberculosis (the bacteria which causes TB). By identifying the wells with bacterial growth, this project will help researchers determine which antibiotics are effective at killing each specific strain of TB.
As for 2018, I’m excited about the prospect of sequencing all 500,000 volunteers in UK Biobank, something I wrote about in the Observer on New Year’s Eve.
Ones to watch
I always enjoy animations that reveal how proteins function, and this one of Dicer, based on results obtained by cryo-electron microscopy, is particularly good.
In my next blogpost I’ll be concentrating on the third pillar of our science strategy. Meanwhile, any thoughts on career obstacles, sharing science facilities or great science I’ve missed, do let me know via Facebook, Twitter or LinkedIn.