Big Data Approaches

Gupta, R. (Organizer), Andrea Califano (Panel member), Thorkild I.A. Sørensen (Panel member), Atul Butte (Panel member), Elaine Mardis (Panel member)

    Activity: Attending an eventParticipating in or organising a conference


    Large-scale data generation, including whole-genome sequencing (WGS), transcriptomics, proteomics,
    metabolomics, microbiome, flow and mass cytometry, and imaging is rapidly becoming more accessible in clinical
    setups; but methods to integrate and infer from data have not kept up, and have not matured or attained
    credibility for adoption in clinical communities. Big Data and associated methodologies for subgrouping cohorts or
    addressing individual patients have the potential to impact treatment options and ultimately improve patient
    health through more targeted and timely therapy1–3. However, there is a need for further methods development,
    prototyping on diseases with detailed phenotyping, education and communication before such methods can
    seriously influence clinical decisions.
    Methods to predict disease progression and treatment outcomes—efficacy and toxicities— should ideally
    incorporate large-scale patient data of mixed types, electronic health records (EHRs) and registries, and prior
    biological knowledge from literature and cell-lines. Several complex diseases are treated through longitudinal
    protocols, some over several years, where data from patient visits provides an opportunity to adjust treatment
    and predict late effects. Such data is rarely used in comprehensive models, but current emerging methods in in
    data integration, systems biology, machine learning (ML), Bayesian methodologies offer ideal opportunities to
    benefit from the wealth of patient data that is currently available.
    The symposium will explore these themes through current projects that utilise Big Data methodologies in
    addressing treatment. The aim of the day will be to focus discussions on meaningful individual patient benefit,
    and outlining challenges that need to be addressed to make personalised medicine a reality. Addressing such
    challenges takes significant collaboration as well as educational programmes across data scientists, clinicians,
    clinical epidemiologists and biostatisticians. It would be our aim to attract a mixed audience, and foster
    challenging discussions – that would lead to new collaborations and opportunities.
    Period15 May 2018
    Event typeConference
    LocationHellerup, Denmark
    Degree of RecognitionInternational