Wellcome Mental Health Data Prize
The Wellcome Data Prize in mental health will support collaborative approaches to research into anxiety and depression in young people. Teams in the UK and South Africa will explore existing data to find new insights and build digital tools that enable future research.
Prize at a glance
- Where the host organisation of the lead applicant is based:
- South Africa, UK
- Level of funding:
There are three phases to the prize, each six months long and with different levels of funding and support on offer:
- Discovery phase: 10 teams will be selected to receive £40,000 of funding
- Prototyping phase: 5 teams will be selected to receive £100,000 of funding
- Sustainability phase: £500,000 will be allocated across 3 winning teams
- Duration of funding:
About this prize
The Prize aims to generate tangible and scalable outputs that support the mental health research community. The overarching question for the Data Prize is:
What are the ‘active ingredients’ that make a difference in preventing, treating, and managing anxiety and depression in young people? What works, for whom, in what contexts, and why?
'Active ingredients' are those aspects most likely to make a difference in preventing, treating or managing mental health difficulties. This means they:
- drive clinical effect
- are conceptually well-defined, and
- link to specific, hypothesised mechanisms of action.
Since 2020, Wellcome has been commissioning research teams from across the world to review the evidence for different active ingredients deemed to help prevent, treat, and manage anxiety and/or depression in 14 to 24-year-olds globally.
So far, over 50 research teams have been commissioned to review 46 active ingredients. These active ingredients can be broadly categorised by their focus of change into six groups:
- behaviours and activities
- beliefs and knowledge
- brain/body functions
- cognitive and attentional skills
- human connections
- socioeconomic factors
Reviews into active ingredients have aimed to improve understanding of:
- whether existing evidence shows that their active ingredient is effective among 14 to 24-year-olds
- whether there are subgroups or contexts in which their ingredient is particularly effective or ineffective
- the mechanisms of action underpinning the efficacy of their active ingredient.
It is important to stress that the 46 active ingredients focussed on so far do not represent an exhaustive list. Applicants to the prize are encouraged to propose research that expands the scope of the existing research, for example, by identifying new active ingredients. Additionally, the age range of children and young people of interest to this prize (0-30-year-olds) is wider than in the reviews conducted to-date, providing further opportunities for innovation in research.
There are three phases to the prize, with only a subset of teams from each phase progressing to the next:
1. Discovery Phase (10 teams)
- Participants will be directed to exemplar longitudinal datasets from the UK and South Africa, and can bring their own data where relevant.
- Teams will produce an analysis of existing data to answer a research question.
- Teams will be supported with data acquisition, and refining use cases (further examination of how, why and where the digital tool they develop could be used) as well as lived experience involvement approaches.
2. Prototyping Phase (5 teams)
- Teams will create a prototype of a digital tool, and they will also refine and disseminate their research from the discovery phase.
- Teams will be supported with further developing use cases, lived experience involvement approaches and workshops on agile principles and prototyping.
3. Sustainability Phase (3 teams)
- Support includes data science and product design expertise, support on planning user testing and lived experience involvement approaches and continuous feedback sessions.
Eligibility and suitability
Scope of proposals
Applicants must submit a research question in an area of their interest. Then, the successful awardees will explore existing datasets to help answer that question while developing their analysis tool.
Proposals will be evaluated on impact, innovation and feasibility. Key considerations also include the breadth of skills within teams and how teams plan to involve lived experience of youth anxiety and depression throughout their work.
Analysis does not need to be limited to the 46 active ingredients where we have previously commissioned research and can be on aspects of any intervention most likely to be contributing to making the difference in preventing, treating, or managing ongoing mental health difficulties.
Primary research is out of scope and proposals should not intend to perform primary data collection.
By the end of the award, teams will have produced:
- a digital tool to facilitate data analysis in the mental health research community, and
- the analysis of the datasets they explored.
What do we mean by digital tools?
Tools will build on insights formed in the Discovery phase.
The primary users of these tools will be mental health researchers with the purpose of enabling research into active ingredients, those aspects of an intervention that drive a clinical effect.
Many types of tools could be developed in the prize and we will accept prototypes as an output.
Below are the types of tools that are considered in scope for the Prize.
- Tools helping researchers to perform data analysis, for example a tool that can identify clusters of individuals who respond to specific interventions or active ingredients, or a tool that determines which factors can predict relapses.
- Tools used to replicate data analysis, by making available the mechanism by which data is translated into research insights, for example, a new application of a machine learning algorithm that can be used on other datasets.
- Tools that share the insights from data analysis, in a format accessible and digestible for multiple audiences, for example, a triaging tool for researchers that can highlight the active ingredients that work for different groups of young people.
- Tools that facilitate data analysis by addressing barriers to conducting research, for example, tools that support data cleaning and manipulation or automatically extract relevant data from longitudinal datasets.
Please note this list is not exhaustive and serves only as a guideline for what may be developed.
Tools considered out of scope would be:
- Tools based on insights from data but not directly linked to data itself, for example, analysis shows that certain interventions are effective leading to a tool connecting individuals that use that intervention.
- Apps aimed at individuals, for example, condition management, mood tracking, sleep/exercise monitoring apps or a digital tool that supports people directly with navigating existing services for themselves.
Teams do not necessarily need to pursue the same digital tool as proposed in the application, particularly if the need for the tool evolves as a result of the analysis performed. We encourage teams to incorporate their learnings from the discovery phase into the digital tool they ultimately develop.
Intellectual property (IP)
The IP developed with Wellcome’s funding should be owned by the organisation with principal responsibility for administering the grant. That notwithstanding, the IP should be made available publicly on an appropriate open-source software licence in line with Wellcome’s open access policy.
A key aim of this prize is to build a multidisciplinary mental health data community by bringing together people with mental health research backgrounds and data expertise. By mental health science we mean any discipline that uses evidence in rigorous and transparent ways, whether based on observation or experimentation. This could include:
- disciplines within the humanities, social sciences and computer sciences, among others.
We also welcome early career researchers.
We understand that groups may want to apply but feel they require additional skillsets, and we will actively support the formation of teams where possible by connecting expertise to those with similar interests and aims across organisations.
For applicants who do not yet have a full multi-disciplinary team, Social Finance will be coordinating opportunities to form connections with other interested organisations. Please sign up to the mailing list to find out more about these events and join the Slack channel to meet other people interested in forming a team.
Please see the eligibility criteria below for further information on who can apply.
- must be based in the UK or South Africa.
- may be at any career stage but must have a permanent, open-ended or long-term rolling contract, or the guarantee of one for the full duration of the award.
Lead applicants must be based at an eligible host organisation that can sign up to our grant conditions.
Only the lead applicant from each team needs to be based in the UK or South Africa
Lead applicant organisations should be either:
- a higher education institution
- a research institute
- a non-academic healthcare organisation
- a not-for-profit organisation.
- can be at any career stage and based anywhere in the world, apart from mainland China.
- can be self-employed, e.g. a freelance data scientist (as well as employed by the usual range of host organisations below).
Co-applicants must be based at an eligible organisation that can sign up to our grant conditions.
Co-applicants organisations should be either:
- higher education institution
- research institute
- non-academic healthcare organisation
- not-for-profit organisation
- commercial organisation.
- freelancers – such as data scientists
All applicants must agree to our standard grant conditions.
All applicants must make a significant and essential contribution to the work.
Teams can propose their own datasets as research data, with the potential gain of opening them for analysis from a range of expertise after the prize.
Participating teams can bring their own data, so long as they meet the following criteria:
- Prize teams led by UK researchers must conduct research on datasets with UK cohorts. Prize teams led by South African researchers must conduct research on datasets with South African cohorts.
- Datasets must include data about respondents under 30 years of age (child / young person respondents) to enable research focused on those respondents.
- Datasets must be longitudinal, allowing for the investigation of temporal effects, with at least three waves of data capture for child / young person respondents.
- Datasets must include data from at least one age-appropriate measure of depression and/or anxiety (or their symptoms), used in child / young person respondents across multiple waves of longitudinal data collection. Proxy measures of these disorders or symptoms are permitted but should be well-justified.
- Datasets must include data suitable for exploring active ingredients of depression and/or anxiety in children and young people.
In addition to this, a selection of longitudinal datasets from the UK and South Africa will be proposed as appropriate ‘exemplar’ research data to participants, with streamlined access to these datasets facilitated where appropriate.
These exemplar datasets are longitudinal research studies and have been selected as they contain rich data that has the potential to provide new insights on active ingredients. Examples of these datasets are the Avon Longitudinal Study of Parents and Children (ALSPAC) and the Millennium Cohort Study.
Within their proposals, participants must be able to demonstrate both that their dataset is:
- appropriate for their research question and/or tool
- and that they have gained appropriate access to use this data within the prize.
If applicants have access to a dataset that does not meet all of the criteria above but provides other added value to address the research question set out, please contact us at firstname.lastname@example.org so we can assess on a case-by-case basis.
Teams that bring their own data must be prepared to share the insights they have found along with relevant data, in line with Wellcome’s open access policy.
When submitting an application, all teams must ensure that they include the following information relating to the data they will use:
- The provenance of the data
- Data fields that will be used for research
- A cohort of study members that will be included in the research
- A timeframe of study data that will be included in the research
- An outline of why the chosen data is appropriate for addressing the research question
- Approval to access the data from the relevant data controller, if required
- Confirmation that the use of the data aligns with the legitimate basis for its collection and use
- Organisational data protection policies for safeguarding all data
- An outline of how the project will comply with local privacy laws and, where relevant, access protocols for restricted datasets
- Mitigations against any risks that research outputs may be misused for harmful purposes
- Steps that will be taken to make research outputs publicly available, including:
- analytical dataset
- source code for analysis and digital tools, along with sufficient data to enable others to understand, test, run and re-use these
- research papers.
Evaluating dataset choice
As part of the evaluation process, the evaluation panel will be assessing the suitability of the datasets that applicants propose against these criteria. This applies to teams who intend to use one of the exemplar datasets as well as to teams who intend to propose an alternative dataset.
Prize team applicants must be able to evidence, with the submission of their prize application, confirmation from the relevant data controller that they are permitted to access the data required to pursue their research goals.
In some cases, it may not be possible for teams to gain approval to access highly sensitive data before the application deadline. In these cases, we expect:
- applications to include as much evidence as possible outlining the progress they have made with data holders to access these data and a likely timeframe for full access.
- teams to outline the adjustments they will make to their research proposal in the event they are denied access to sensitive data and the impact this will have on the outcomes of their research.
Applicants should also demonstrate why their choice of dataset is particularly suited to their research question and project goals.
Please be specific about the timeframe, cohort and data fields to be used and how your choice of these is appropriate for your research question.
Expectations for data ethics
As part of the evaluation of data prize applications, we will be evaluating the safeguards that teams have in place to ensure the ethical use of any data they will use. Prize teams must detail in their application the provisions they will have in place for ensuring:
- that their research question and use and any sharing of the data aligns with the provenance of the data and the legitimate basis for its collection
- clear protocols for the safeguarding of all data and in particular sensitive data
- compliance with local data privacy laws and, where relevant, compliance with any access protocols in relation to restricted access datasets or shared data platforms
- mitigations against the risk that their research may generate outcomes that could be misused for harmful purposes
- In all instances, organisational data protection policies must be included as an attachment to proposals to allow for a robust assessment of the safeguarding procedures applied to the proposed dataset/s.
Expectations for open science
Wellcome are committed to conducting mental health research in an open way and we invite prize participants to follow and, where possible, advance methods and ways of working that facilitate community-led research.
Where the data are concerned, we would expect all teams to follow these principles:
- Where relevant, appropriate and privacy-preserving, analytical datasets, along with variables created and transformed during research, should be made publicly available.
- All source code (for research and tool development) must be licensed under an OSI-compliant licence and must be hosted on a freely-available and publicly accessible collaboration repository, such as GitHub, GitLab or BitBucket.
- If any software (including any software tool) depends on or processes data then sufficient data must be made freely available to enable third parties to understand, test, run, and re-use the relevant software. This could be a small sub-set of data embedded into the code repository, or a link to data on a publicly-accessible repository.
- Publications resulting from prize research should be published in open access journals, wherever possible.
Please make reference to the Wellcome Open access policy for further details.
We actively encourage prize participants to engage in a research community comprised of fellow prize teams. This includes sharing insights and outputs of research and collaborating on data science challenges specific to analysing longitudinal datasets. This will benefit from the development of open science methods that enable such sharing without disclosure of sensitive data about study respondents, such as the generation of synthetic datasets.
We have consulted with several data holders in the UK and South Africa to build an understanding of the data that we consider to be appropriate for the prize. We have identified several exemplar datasets that meet all of the criteria above. The data holders of each of these datasets have been consulted on the most effective way to access their data, with details documented in the 'Exemplar datasets' PDF documents below.
We welcome applications from teams that propose to conduct their research on alternative datasets, provided they meet the criteria above. Applications proposing the use of such datasets will not be disadvantaged for selection.
When making an application to a data holder for access to a dataset, please use the prefix “Wellcome MH data prize” to your project title.
For each of the feasibility criteria, applicants will receive a pass or fail score. All the feasibility criteria need to receive a pass score as a prerequisite for selection.
Pass/fail (PF) feasibility criteria
- Where relevant, the team has approval and has met all the appropriate requirements and standards to access the data required for their research from the relevant data controller.
- The choice of dataset is appropriate to address the research question and aligns with the legitimate basis for collection and use.
- The team demonstrates that they have adequate organisational data protocols to safeguard sensitive data that are consistent with applicable local and regional regulations.
- The team has considered research ethics and data privacy and has a plan to mitigate against potential risks.
- The team has provided a suitable research methodology to address their research question.
- The team has considered time and resource implications in their research plan and has provided evidence that they will be able to make meaningful progress within the initial 6-month period.
- The team shows evidence of sufficient time and resources and has a reasonable plan in place on how the initial prototype will be achieved during the 6-month prototyping phase.
- The team have a plan that will involve the wider community in the project, by embedding mental health lived experience and/or sharing insights/methods with the research/data science community.
- Team members have the appropriate skills and experience to be able to answer the research question.
- Team members have the appropriate skills and experience to be able to develop the proposed tool.
- Team members have the appropriate skills and experience to be able to deliver an involvement plan.
- Applications that pass the feasibility criteria will receive a score between 1 and 5 for both impact and innovation across each of the three criteria:
- The research question advances understanding of 'active ingredients' of effective interventions for anxiety and/or depression in young people, including what helps, for whom, in what contexts, and why (for example, mechanisms underpinning effectiveness). Outputs will point to how insights gained may impact young people experiencing anxiety and/or depression, having the potential to significantly impact either a wide range of people experiencing anxiety and/or depression, or a group who are not commonly represented within mental health research.
- Dataset variables and analytical methods are clearly articulated and well-suited to the research question.
Note: this criteria is designed to allow for a more nuanced evaluation of the proposals above the baseline feasibility criteria 1 and 5 below.
- As per the Wellcome open access policy, the team provides a convincing summary approach of how they plan to ensure that all outputs, including publications, analytical datasets along with variables created and transformed during research, will be made publicly available.
- The team demonstrates excellent understanding of the existing research connected with the chosen active ingredient and demonstrates that their research question is novel and has the potential for innovation in mental health science (for example, it could point the way towards improving existing early interventions for anxiety and/or depression, or developing new ones).
- The research methodology is innovative and has the potential for wider use in other research applications.
Note: it is not expected that the team must innovate in their methodology if this is not appropriate for their research.
- The proposed tool will aid/enable future research into youth anxiety and/or depression, and/or the active ingredients approach in mental health research.
- The proposed tool has applications within mental health research beyond the scope of this data prize. Note: it is not expected that the team must demonstrate this.
- Proposal details approaches that the team will take to make tools available for wider use, including using publicly accessible code repositories, providing sufficient data and testing to enable third parties to understand, test, run and re-use software.
- The approach to the prototype phase follows sensibly from discovery phase with consistency in planning
- The proposed tool has the potential to provide a novel way of addressing data and/or research needs within mental health research into active ingredients.
- There is evidence that the multidisciplinary composition of the team is crucial to the success and impact of the proposed project.
- The team shows a plan for how they will be working collaboratively to share and synergise their complementary expertise throughout the process (avoiding siloed work).
- Where relevant, the involvement plan will positively impact people with lived experience working on the project.
- The team has a plan to share work/embed research beyond the mental health research community.
- All key partners demonstrate a clear sense of their roles and have a consistent way of talking about the project aims and ways of working.
- Team formation brings together organisations that have not traditionally worked together.
- The proposed team demonstrates one or more of the following:
- a new, cross-disciplinary way of approaching research into active ingredients of depression and/or anxiety in young people
- involvement with people with lived experience
- the plan for working incorporates reflexive practice (awareness of influence a researcher has on the people/topic being studied), and significantly addresses the traditional power structure of a mental health research team
- Team formation and involvement plans:
- promotes diversity and inclusion in terms of skills and background
- takes steps to be representative of the population involved in the research
- The involvement plan is innovative in the way the team intends to adapt to or have feedback strongly impact their research and tool production.
- The team approach has the potential to advance and share awareness of best practice use of data in the mental health research community.
Each of these criteria will be equally weighted. More details are in the 'Useful documents and links' section. The best scoring teams will be invited to pitch.
- Pitches will provide an opportunity for applicants to provide more depth to their proposal and for the evaluation panel to ask questions.
- After the pitch, the 10 prize teams will be selected.
Applications will be reviewed by the following panel:
- Professor Ann John (Chair), Co-Director DATAMIND, Swansea University
- Stefaan G. Verhulst, Co-founder and Chief Research and Development Officer, The Governance Laboratory at New York University
- Professor Andrea Mechelli, Institute of Psychiatry, Psychology & Neuroscience, King's College London
- Professor Crick Lund, Professor of Global Mental Health, Centre for Global Mental Health, King’s College London and Honorary Professor, Alan J Fisher Centre for Public Mental Health, Department of Psychiatry and Mental Health, University of Cape Town
- Professor Katherine Sorsdahl, Co-Director, Alan J Fisher Centre for Public Mental Health Department of Psychiatry and Mental Health, University of Cape Town
- Jodie Crooks, Research Assistant, University of Leeds (Lived Experience Expert)
- Jaimie Lee Davids, Training Officer, Waves for Change (Lived Experience Expert)
- Alessandra Fassio, Advocacy and Relations Manager – Data for Children Collaborative with UNICEF
What we offer
There are three phases to the prize, each six months long and with different levels of funding and support on offer:
- Discovery Phase: an initial 10 teams will be selected to receive £40,000 of funding.
- Prototyping Phase: five teams from the Discovery Phase will receive £100,000.
- Sustainability Phase: three winning teams will receive £500,000.
The amounts listed are per team. Throughout the prize, teams will be supported with data access, analysis and use case development as well as support on planning lived experience involvement approaches.
There will also be regular group activities such as problem-solving sessions and theory of change workshops.
How to apply
Teams will apply with a written proposal and shortlisted teams will be asked to present their proposals to a panel.
1. Before you apply
Make sure you read everything on this page.
2. Submit your application
- View the sample application form [PDF 194KB] to understand what you will need to include in your application.
- Complete your application on Wellcomedataprize.submit.com
Applications will be reviewed based on their feasibility, impact and innovation. For details please see the evaluation criteria.
4. Selection process
Initial Application Review
Social Finance will review applications for eligibility and appropriate data access for teams that wish to use their own datasets.
Panel Application Review
The selection panel and data ethics advisor will review the relevant sections of each application based on their area of expertise. Applications will be assessed using the evaluation criteria and up to twenty teams will be shortlisted.
Shortlisted teams will be asked to present to the selection panel who will use the evaluation criteria to select the final ten teams.
Final funding decisions will be made by Wellcome following the recommendation from the judging panel. Please note we are not able to provide feedback to unsuccessful applicants
5. Progression to Prototyping phase
Teams will be evaluated on their progress and performance in the Discovery Phase and five teams will be selected to take part in the Prototyping phase. Evaluation will be carried out by a selection panel chosen for their expertise in relevant fields.
If you are disabled or have a long-term health condition, we can support you with the application process.
You must submit your application by 17:00 (GMT/BST) on the deadline day. We do not accept late applications.
Open to New Applications
- 5 June 2022
- 24 June 2022
Confirmation of shortlisted teams
- 5, 6 or 7 July 2022
Teams will pitch on the afternoon of one of the following days
- Week commencing 18 July 2022
Selection of 10 teams for Discovery phase
- 8 August 2022
First day of the Discovery phase