Enabling the big data revolution through skills training

Tremendous amounts of data are produced every day in the life sciences and biomedicine and made available to the wider biological research community in the form of databases, data records and digitised information. The heterogeneity of these data, and at the same time, the potential amount of information that they can give us is astoundingly complex. We are at a crucial moment in biomedical research for which the effort of multiple disciplines in the quantitative and in the biomedical sciences have to come together to rationalise, quantify and extract the essential information content that can be translated into practical use in the medical context.
We present a flexible training programme to facilitate the understanding of the data landscape that is populating modern medicine. More accurate and informed diagnoses will be possible if all the involved parties are able to extract useful information from patient data records and if these are efficiently integrated with genetic and analytical investigations with the aim of designing personalised therapies. The program is aimed at a large set of trainees: from medical practitioners, clinicians, scientists, companies and workers in the health sector. The courses will focus on skills training of different complexity that can be assembled in a personalised modular fashion.
The flexibility is in the opportunity to pick and mix courses to generate learning curricula of different depth levels that can be started at any point.
The offered courses will range from data exploration, integration and manipulation to more in depth analyses via computational statistical and artificial intelligence (AI) based methods. The trainees will have the opportunity to participate in the assembly of computational pipelines to analyse the data, and to bring to the table their own data for collaborative analyses.

We have designed the training programme centred around three pillars (workstreams) that we believe are among our strengths in terms of training expertise, data collection and method development: Health Data Science exploring electronic data records (WS1); 'Omics harnessing genetics and molecular data collected in online databases (WS2); Artificial Intelligence focusing data image analysis and understanding AI through practical applications (WS3). The cross-talk between these areas of research is only at the beginning, this programme should facilitate collaborative efforts in identifying and overcoming the barriers for effective integration and translation across disciplines.
The programme will engage the supporters, the patients and the public in workshops led by the participants sharing their learning experience and feedback suggestions for the structure and contents of the thought material.

The program has three levels of governance: A) a management committee of PI, co-I's and WS leads; B) a stakeholder committee of representatives from academic research including ECRs, mid-career and senior leaders, clinical trainees and clinical academics and industry-based trainees; C) an advisory group with representatives invited from the funder and project partners to feedback information in a loop from which the program will continuously learn and improve.

Biomedicine research is at the crossroads of a transformative change if an effective use of all the data gathered in the field can be effectively rationalised and integrated. At present, data are stored in separate silos that are specific to a research field and its culture. It is becoming apparent that cross-disciplinary training encompassing these cultures is the best way forward to connect researchers and to integrate data from those silos.
Non-standardised data formats and lack of generic data pipelines are hampering effective integration and translation across disciplines and therefore limiting scientific cross-communication.
We present a training programme with the ambitious task of bridging the gap between real-world clinical care data, genetics, molecular and imaging data with computational statistical modelling and Artificial Intelligence. We believe we have excellent research portfolio in these areas and proven success in a number of active teaching activities, but we see the opportunity of integrating these in a truly interdisciplinary programme that is flexible and modular and allows for the design of personalised training pathways that can be initiated at any time.
The programme will have a circular structure with no specific entry point and with the possibility of pick and mix courses from the three areas: Health Data Science exploring electronic data records (WS1); 'Omics harnessing genetics and molecular data collected in online databases (WS2); Artificial Intelligence focusing data image analysis and understanding AI through practical applications (WS3).
The courses will cover the understanding of online 'omics data resources, harnessing the heterogeneity of health informatics and imaging records by interacting with the data at different levels: From the implementation of simple pipelines to analyse data, to the mastering of statistical analyses, to clinical prediction modelling, machine learning and artificial intelligence data analyses.