All posts by Office of Scholarly Communication

Sustaining open research resources – a funder perspective

Uncategorized, , ,

This is the second in a series of three blog posts which set out the perspectives of researchers, funders and universities on support for open resources. The first was Open Resources, who should pay? In this post, David Carr from the Open Research team at the Wellcome Trust provides the view of a research funder on the challenges of developing and sustaining the key infrastructures needed to enable open research.

As a global research foundation, Wellcome is dedicated to ensuring that the outputs of the research we fund – including articles, data, software and materials – can be accessed and used in ways that maximise the benefits to health and society.  For many years, we have been a passionate advocate of open access to publications and data sharing.

I am part of a new team at Wellcome which is seeking to build upon the leadership role we have taken in enabling access to research outputs.  Our key priorities include:

  • developing novel platforms and tools to support researchers in sharing their research – such as the Wellcome Open Research publishing platform which we launched last year;
  • supporting pioneering projects, tools and experiments in open research, building on the Open Science Prize which with the NIH and Howard Hughes Medical Institute;
  • developing our policies and practices as a funder to support and incentivise open research.

We are delighted to be working with the Office of Scholarly Communication on the Open Research Pilot Project, where we will work with four Wellcome-funded research groups at Cambridge to support them in making their research outputs open.  The pilot will explore the opportunities and challenges, and how platforms such as Wellcome Open Research can facilitate output sharing.

Realising the long-term value of research outputs will depend critically upon developing the infrastructures to preserve, access, combine and re-use outputs for as long as their value persists.  At present, many disciplines lack recognised community repositories and, where they do exist, many cannot rely on stable long-term funding.  How are we as a funder thinking about this issue?

Meeting the costs of outputs sharing

In July 2017, Wellcome published a new policy on managing and sharing data, software and materials.  This replaced our long-standing policy on data management and sharing – extending our requirements for research data to also cover original software and materials (such as antibodies, cell lines and reagents).  Rather than ask for a data management plan, applicants are now asked to provide an outputs management plan setting out how they will maximise the value of their research outputs more broadly.

Wellcome commits to meet the costs of these plans as an integral part of the grant, and provides guidance on the costs that funding applicants should consider.  We recognise, however, that many research outputs will continue to have value long after the funding period comes to an end.  Further, while it not appropriate to make all research data open indefinitely, researchers are expected to retain data underlying publications for at least ten years (a requirement which was recently formalised in the UK Concordat on Open Research Data).  We must accept that preserving and making these outputs available into the future carries an ongoing cost.

Some disciplines have existing subject-area repositories which store, curate and provide access to data and other outputs on behalf of the communities they serve.  Our expectation, made more explicit in our new policy, is that researchers should deposit their outputs in these repositories wherever they exist.  If no recognised subject-area repository is available, we encourage researchers to consider using generalist repositories – such as Dryad, FigShare and Zenodo – or if not, to use institutional repositories.  Looking ahead, we may consider developing an orphan repository to house Wellcome-funded research data which has no other obvious home.

Recognising the key importance of this infrastructure, Wellcome provides significant grant funding to repositories, databases and other community resources.  As of July 2016, Wellcome had active grants totalling £80 million to support major data resources.  We have also invested many millions more in major cohort and longitudinal studies, such as UK Biobank and ALSPAC.  We provide such support through our Biomedical Resource and Technology Development scheme, and have provided additional major awards over the years to support key resources, such as PDB-Europe, Ensembl and the Open Microscopy Environment.

While our funding for these resources is not open-ended and subject to review, we have been conscious for some time that the reliance of key community resources on grant funding (typically of three to five years’ duration) can create significant challenges, hindering their ability to plan for the long-term and retain staff.  As we develop our work on Open Research, we are keen to explore ways in which we adapt our approach to help put key infrastructures on a more sustainable footing, but this is a far from straightforward challenge.

Gaining the perspectives of resource providers

In order to better understand the issues, we did some initial work earlier this year to canvas the views of those we support.  We conducted semi-structured interviews with leaders of 10 resources in receipt of Wellcome funding – six database and software resources, three cohort resources and one materials stock centre – to explore their current funding, long-term sustainability plans and thoughts on the wider funding and policy landscape.

We gathered a wealth of insights through these conversations, and several key themes emerged:

  • All of the resources were clear that they would continue to be dependent on support from Wellcome and/or other funders for the long-term.
  • While cohort studies (which provide managed access to data) can operate cost recovery models to transfer some of the cost of accessing data onto users, such models were not appropriate for data and software resources who commit to open and unrestricted access.
  • Several resources had additional revenue-generation routes – including collaborations with commercial entities– and these had delivered benefits in enhancing their resources.  However, the level of income was usually relatively modest in terms of the total cost of sustaining the resource. Commitments to openness could also limit the extent to which such arrangements were feasible.
  • Diversification of funding sources can give greater assurance and reduce reliance on single funders, but can bring an additional burden.  There was felt to be a need for better coordination between funders where they co-fund resources.  Europe PMC, which has 27 partner funders but is managed through a single grant is a model which could be considered.
  • Several of the resources were actively engaged in collaborations with other resources internationally that house related data – it was felt that funders could help further facilitate such partnerships.

We are considering how Wellcome might develop its funding approaches in light of these findings.  As an initial outcome, we plan to develop guidance for our funded researchers on key issues to consider in relation to sustainability.  We are already working actively with other funders to facilitate co-funding and make decisions as streamlined as possible, and wish to explore how we join forces in the future in developing our broader approaches for funding open resources.

Coordinating our efforts

There is growing recognition of the crucial need for funders and wider research community to work together develop and sustain research data infrastructure.  As the first blog in this series highlighted, the scientific enterprise is global and this is an issue which must be addressed international level.

In the life sciences, the ELIXIR and US BD2K initiatives have sought to develop coordinated approaches for supporting key resources and, more recently, the European Open Science Cloud initiative has developed a bold vision for a cloud-based infrastructure to store, share and re-use data across borders and disciplines.

Building on this momentum, the Human Frontiers Science Programme convened an international workshop last November to bring together data resources and major funders in the life sciences.  This resulted in a call for action (reported in Nature) to coordinate efforts to ensure long-term sustainability of key resources, whilst supporting resources in providing access at no charge to users.  The group proposed an international mechanism to prioritise core data resources of global importance, building on the work undertaken by ELIXIR to define criteria for such resources.  It was proposed national funders could potentially then contribute a set proportion of their overall funding (with initial proposals suggesting around 1.5 to 2 per cent) to support these core data resources.

Grasping the nettle

Public and charitable funders are acutely aware that many of the core repositories and resources needed to make research outputs discoverable and useable will continue to rely on our long-term funding support.  There is clear realisation that a reliance on traditional competitive grant funding is not the ideal route through which to support these key resources in a sustainable manner.

But no one yet has a perfect solution and no funder will take on this burden alone.  Aligning global funders and developing joint funding models of the type described above will be far from straightforward, but hopefully we can work towards a more coordinated international approach.  If we are to realise the incredible potential of open research, it’s a challenge we must address

Published 26 July 2017
Written by David Carr, Wellcome Trust (d.carr@wellcome.ac.uk)

Creative Commons License

Whose money is it anyway? Managing offset agreements

Uncategorized, , , , , ,

Sometimes an innocent question can blow up a huge discussion, and this is what happened recently at an RCUK OA Practitioner’s Group meeting when I asked what was appropriate for institutions to do when managing money they receive as refunds from publishers through offsetting arrangements.

When an institution pays for an article processing charge (APC) in a hybrid journal, it is doing so in addition to the existing subscription. This is generally referred to as ‘double dipping’.  I have written extensively about the issues with hybrid in the past, but here, I’d like to discuss the management of offset agreements.

Offset agreements are a compensation by a publisher to an institution for the extra money they are putting into the system through payment of APCs. Most large publishers have some sort of offset agreement for institutions in the UK which are negotiated by Jisc, based on the principles for offset agreements. (There is one significant publisher which is an exception because it insists there is no need for an offset agreement because it does not double dip.)

Offset agreements are not equal

While offset agreements are negotiated nationally, there is no obligation for any institution  to sign up to them. Cambridge makes the decision to sign up to an offset agreement or not through a standard calculation. If we are spending RCUK and COAF funds on the offset it must show benefit to the funds first. If the numbers demonstrate that by signing up to (and sometimes investing in) the agreement, the funds will be better off at the end of the year then we sign. The fact this agreement may have a broader benefit to the wider University is a secondary consideration. The OSC has a publisher and agreements webpage listing the agreements Cambridge is signed up to.

In a fit of spectacular inefficiency, all offsets work slightly differently. Here’s a run down of different types:

  • In some instances we have a melding of the costs into one payment and there are no transactions for open access. The Springer Compact is an example of this. At Cambridge we have split the cost of this deal between the subscription spend the previous year with the top up being made by our funds from RCUK and COAF in proportion to the amount we publish between these two funders with Springer.
  • Other offsets are internal – where the money does not leave the publisher’s system. The Wiley OA Agreement is this type. By signing up we receive a 25% discount on each APC that is managed through their dashboard. We also receive a 50% discount in a given year based on the number of APCs we bought the previous year. This money is calculated at the beginning of the year and the ‘money’ is put into a ‘fund’ held by Wiley. The APC payments for future articles can be made out of this credit. It is is bit like a betting app – you can’t get the money out without some difficulty, you can only ‘reinvest’ it
  • There is a different kind of internal offset where the calculation is made up front based on how much you spent the previous year on APCs. These manifest as a discount on each APC paid. Taylor and Francis’ offset works this way which is a bit of a hassle because you still have to process each APC regardless of whether you spend $2000 or $200 on it. But again there is no extra money anywhere in this equation because the discount is applied before the invoice is issued. 
  • A different kind of arrangement relates more to fully open access journals. These include a membership where you get a discount on APCs for being a member. Sometimes there is a payment associated with this (BMC for example, which for an upfront membership you can get 15% discount), and others where there is no payment (MDPI – 10% discount for now). Alternatively you can ‘buy’ membership for researchers in exchange for the right to publish for free (PeerJ).
  • The last type of offset is the most straightforward – where the institution gets a cheque back based on the extra spend on APCs over the subscription. Currently IoP is the only publisher with whom Cambridge has this type of agreement.

Managing offset refunds

When Cambridge received its first IoP cheque in 2015 there were questions about what we could or could not do with it. The Open Access Project Board discussed the issue and decided that the money needed to remain within the context of open access. Suggestions included paying our Platinum membership of arXiv.org with it, because this would be supporting open access.

The minutes from the meeting on 31 March 2015 noted: “Any funds returned from publishers as part of deals to offset the cost of article processing charges should be retained for the payment of open access costs, but ring-fenced from the block grants and kept available for emergency uses under the supervision of the Project Board.” We have since twice used this money to pay for fully open access journal APCs when our block grant funds were low. 

Whose money is it anyway?

When the issue of offset refunds and what institutions were doing with it was raised at a recent RCUK OA Practitioners Group meeting it became clear that practices vary considerably from institution to institution. One of the points of discussion was whether it would be appropriate to use this money to support subscriptions. The general (strong) sentiment from RCUK was that this would not be within the spirit, and indeed against the principles, of the RCUK policy.

I subsequently sent a request out to a repository discussion list to ask colleagues across the UK what they were doing with this money. To date there have only been a handful of responses.

In one instance with a medium-sized university the IoP money is placed into a small Library fund that is ring-fenced to pay for Open Access in fully Open Access journals only. This fund has the strategic aim to enable a transition to Open Access by supporting new business models and contributing to initiatives such as Knowledge Unlatched, hosting Open Journal Systems, as well as supporting authors to publish in Open Access venues when they have no other source of funding.

A large research institution responded to say they had a specific account set up into which the money was deposited, noting, as did the other respondents, that the financial arrangements of the University would mean that if it were deposited centrally it would never be seen again. This institution noted they were considering using the funds to offset the subscription to IoP in the upcoming year due to a low uptake of the deal.

Another large research institution said the IoP cheques were being ‘saved’ in the subscriptions budget.

Sussex University

In their recent paper “Bringing together the work of subscription and open access specialists: challenges and changes at the University of Sussex” there is a section on how they are managing the offset money. They note: “It seemed a missed opportunity to simply feed it back into the RCUK block grant, but equally inappropriate to use for journal subscriptions or general Library spending”.

The decision was to support APCs for postgraduate researchers (PGRs) who did not have any other access to money for gold open access, and could only be spent on fully open access journals. They noted that this was a welcome opportunity to be able to offer something tangible and helpful in their advocacy dealings with postgraduate researchers.

Only the start of the conversation

This discussion has raised questions about the decision making process for supporting access to the literature.

Subscriptions are paid for at Cambridge through a fund that is not owned by the Library – the fund consists of contributions from all the Schools plus central funds. Representatives of the Schools, Colleges and library staff sit on the Journal Coordination Scheme committee to decide on subscriptions. However decisions about open access memberships and offsets are made by the Office of Scholarly Communication. Given the increased entanglement of these two routes to access the literature, this situation is one the University is aware needs addressing. The Sussex University paper discusses the processes they went through to merge the two decision making bodies.

This is a rich area for investigation – as we move away from subscription-only spend and into joint decision-making between the subscription team and the Open Access team we need to understand what offsets offer and what they mean for the Library. This discussion is just the beginning.

Published 30 June 2017
Written by Dr Danny Kingsley 
Creative Commons License

Open Resources: Who Should Pay?

Uncategorized, , , , , ,

This blog is the first in a series of three which considers the perspectives of researchers, funders and universities in relation to the support for open resources, coordinated and written by Dr Lauren Cadwallader. This post asks the question: What is the responsibility of national funders to research resources that are internationally important?

In January 2017 the Office of Scholarly Communication and Wellcome Trust started an Open Research Pilot Project to try to understand how we could help our researchers work more openly and what barriers they faced with making their work open. One of the issues that is a common theme with the groups that we are working with is the issue of the sustainability of open resources.

The Virtual Fly Brain Example

Let’s take the Connectomics group I am working with for example. They investigate the connections of neurons in fly brains (Drosophila). They produce a lot of data and are committed to sharing this openly. They share their data via the Virtual Fly Brain platform (VFB).

This platform was set up in 2009 by a group of researchers in Cambridge and Edinburgh; some of the VFB team are now also involved in the Connectomics group so there is a close relationship between these projects. The platform was created as a domain-specific location to curate existing data, taken from the literature, on Drosophila neurons and for curating and sharing new data produced by researchers working in this area.

Initially it was set up thanks to a grant from the Biotechnology and Biological Sciences Research Council (BBSRC). After an initial three year grant, the BBSRC declined to fund the database further. One likely reason for this is that the BBSRC resources scheme explicitly favours resources with a large number of UK users. The number of UK researchers who use Drosophila brain image data is relatively small (<10 labs), whereas the number of international researchers who use this data is relatively large, with an estimated 200 labs working on this type of data in other parts of the world.

Subsequently, the Wellcome Trust stepped in with funding for a further three years, due to end in September 2017. Currently it is uncertain whether or not they will fund it in the future. By now, almost eight years after its creation, VFB has become the go-to source for openly available data on Drosophila brain information and images integrated into a queryable platform. No other resource like it exists and no other research group is making moves to curate Drosophila neurobiology data openly. The VFB case raises interesting and important questions about how resources are funded and the future of domain specific open infrastructures.

The status quo

On the one hand funders like the Wellcome Trust, Research Councils UK and National Institutes of Health (NIH) are encouraging researchers to use domain specific repositories for data sharing. Yet on the other, they are acknowledging that the current approaches for these resources are not necessarily sustainable.

A recent review on building and sustaining data infrastructures commissioned by the Wellcome Trust acknowledges that in light of the FAIR principles “it is clear that data is best made available through repositories where aggregation can add most value”, which is arguably in a domain-specific repository. Use of domain-specific repositories allows data to be aggregated with similar data recorded using the same metadata fields.

It is also clear that publishers can influence where data is deposited, with publishers such as Nature Publishing Group, PLOS and F1000 all recommending subject-specific repositories as the first choice place for deposition. If no subject-specific repository is available then unstructured repositories, such as Dryad or figshare are often recommended instead, which complicates infrastructure needs and therefore provisions.

The economic model for supporting data infrastructures is something the Wellcome Trust are considering, with reports recently published by other funding agencies (here, here and here). The Wellcome Trust’s commissioned review noted that project-based funding for data infrastructures in not sustainable in the long term.

However, historically funders have encouraged, and still encourage, the use of domain specific resources, which have been born from project-based funding because of a lack of provision elsewhere. This has created a complex situation – researchers created domain specific data infrastructures using their project funding; these have become the subject norm; funder’s encourage their use, but now don’t have the mechanisms to be able to pledge sustained long-term funding.

National interests?

What is the responsibility of national funders to research resources that are internationally important? Academic research is collaborative. It crosses borders and utilises shared knowledge regardless of where it was generated and this is acknowledged by funders who see the benefits of collaboration. Yet, the strategic goals of funders, such as the BBSRC, are often focused on the national level when it comes to relevance and importance.

On the one hand it is understandable that funders concentrate on national interests – taxpayers’ money goes into the funder’s coffers and therefore they have a responsibility to those taxpayers to ensure that the money is spent on research that benefits the nation.

But, one could argue that international collaboration is in the national interest. The US-based NIH funds resources that are of international importance, including most of the model organism databases and genomic resources, such as the Gene Expression Omnibus. These are highly used by US researchers so one could argue that NIH are acting in the national interest but they are open to researchers all over the world and therefore constitute a resource of international importance.

Wellcome Trust do have a global outlook when it comes to funding, with 21% of their total spend (2015-6) going to projects outside of the UK. Yet, the VFB resource is still vulnerable despite being an internationally important resource.

One of the motivations for the Connectomics group to to participate in the Open Research Pilot is to open a dialogue with the Wellcome Trust about these issues. The Wellcome Trust are committed to strategically investing in Open Research and encourage the use of domain-specific resources. The Connectomics group are interested in how will this strategic investment translate into actual funding decisions now and into the future.

Issues on which researchers would like clarification

All the researchers who are part of the Open Research Pilot have had the opportunity to contribute to questions on open resources sustainability. Posts on the funder’s and University’s perspective will be published as parts 2 and 3 of this blog.

  1. What do you think is the responsibility of national funders towards research resources that are of more international benefit than national?
  2. How do you think the funding landscape will react to the move towards open research in terms of supporting the sustainability of resources used for curating and sharing data?
  3. Researchers are asked to share their data in domain specific resources if they are available. There are 1598 discipline specific repositories listed on re3data.org and each one needs to be supported. How big does a research community need to be to expect support?
  4. What percentage of financial support should be focussed on resources versus primary research?
  5. If funders are reluctant to pay for domain specific resources, is there a need to move to a researcher pays model for data sharing rather than centrally funding resources in some circumstances? Why? How do they envisage this being paid for?
  6. How can we harmonise the approach to sustainable open resources across a global research community? Should we move to centralised infrastructures like the European Open Science Cloud?
  7. More generally how can funders and employers help to incentivise open research (carrot or stick?)
  8. Wellcome often tries to act in a way to bring about change (e.g. open access publishing): Do they envisage that the long term funding of open research (10-20 years from now) will be very different from the situation over e.g. the next 5 years?

Published 23 June 2017
Written by Dr Lauren Cadwallader

Creative Commons License