Seedcorn projects

Academic schools: School of Medicine and the Systems Immunity Research Institute

Researchers: Dr You Zhou, Professor Valerie O’Donnell, Professor Andrew Godkin

You Zhou and team set out to

• develop a common NAFLD module and
• to integrate datasets and model the disease progression.

Both objectives were successfully achieved and, through their seedcorn work, beneficial collaborations were strengthened with Professor Christopher Byrne (University of Southampton) and Professor Hannele Yki-Järvinen (University of Helsinki, Finland).

The team spoke at Workshops, seminars and conferences in Cardiff, Southampton and Helsinki. The first of these saw a Cardiff University medicine student, Nicholas Hodson, addressing a 250 strong audience - the youngest speaker and the only student.

You Zhou said, “With the support from the seedcorn award, we have developed a computational modelling pipeline (developed under support from the Institute). Our work has recently been published in a nature publication group journal with an impact factor at 5.638. I am a co-first author and the seedcorn fund has been highly acknowledged in the paper."

“Furthermore, together with the University of Southampton, we have submitted an application to the NIHR and MRC Efficacy and Mechanism Evaluation Programme fund (£675,864 in total, 11.8% to Cardiff). The application is still under selection, having passed the first round.”

Please contact Dr You Zhou for further information.

Academic schools: School of Physics and Astronomy and School of Psychology

Researchers: Dr Leandro Beltrachini (PI), Dr Andreas Papageorgiou, Dr Cyril Charron, Professor Kevin Murphy, Dr Joseph Whittaker, Professor Matt Griffin

Leandro and team set out to:

• assess the current state-of-the-art in neuroimaging data repositories
• design of an end-to-end Neuroimaging Data Centre (NDC) structure
• pilot study with existing data and
• outline of large grant proposal.

All objectives were achieved. The results of the study are described in detail in a report which has been used to promote the idea of a centre-wide Data Centre and Archive for the Brain Research Imaging Centre (CUBRIC), by providing a comprehensive definition of the structure and operation of such an organisation.

A key element of the work was to bring together a multidisciplinary team from two Schools. Over the duration of the project, this collaboration strengthened and the resulting working group and its aims are expected to continue to function, extending the scope of this current work and looking for further funding. Moreover, the group has developed close links and mutual interests with the Open Science Working Group within CUBRIC, with whom it agreed to have further discussions of this project, as well as future applications.

An additional aim of this work was to develop the concept and demonstrate the benefits of institute-wide, common Data Processing and Archiving within CUBRIC. This study has produced the experience and material necessary to start addressing the challenge. As a result, there are plans to present this work in CUBRIC-wide talks and presentations, with the aim of bringing on board all the different instrument groups in CUBRIC.

Leandro said, “In this seedcorn project, we translated data management practices developed in our previous STFC and UKSA-funded grants in the field of astronomy (six major grants centred in data management between June 2007 and June 2016, totalling more than £1.5m) to the medical imaging data available in the Cardiff University Brain Research Imaging Centre (CUBRIC), enabling researchers worldwide to harness their true potential. The material generated in this project was successfully utilised to get further funding to pursue this challenging enterprise (both from STFC, totaling £185,000)”.

Academic schools: School of Mathematics and School of Medicine

Researchers: Professor Anatoly Zhigljavsky (PI), Professor Valentina Escott-Price

The main aim of the project was to organise a three-day workshop `High-dimensional optimisation and big data’ in the autumn of 2018.

The first day of the workshop (held in London in October) was devoted to machine learning and optimisation in internet advertising. The second and the third days (held on 6 and 7 November in Cardiff) were devoted to theoretical and algorithmic aspects of statistics and machine learning in big data problems including genetics.

Several world-class specialists were invited to Cardiff to give plenary lectures and participate in solving real problems. It was expected that at least two papers in leading statistical journals would be published as a result of the collaborations established in the process of preparation and running the workshop.

Objectives reached

The main objectives of the project were achieved. Among the plenary lecturers at the workshop, several were figures of international standing and very high reputation. The papers which resulted from new collaborations are still in preparation.

As a result of work enabled by the seedcorn fund, collaboration between the School of Mathematics and the School of Medicine has strengthened. A new collaboration between Mathematics and the University of Milan (the group led by Professor Francesco Archetti) has also been established.

In February 2019 Professor Zhigljavsky visited Milan to work with Archetti's group on a variety of projects related to various applications of global optimisation in engineering. At least one joint paper may be expected soon from this collaboration. A discussion about a joint submission to Marie Curie foundation is in progress.

Another useful collaboration has been established with Professor Arthur Gretton from UCL. A collaboration with Oxford group on optimisation headed by Professor Coralia Cartis has also strengthened as a result of the workshop.

The impact of the seedcorn fund is:

• a high-class big data event at Cardiff (attended by more than 100 Cardiff University staff and students from different departments)
• the establishment of new collaborations in the areas of machine learning and big data and strengthening existing collaborations.