Project

Back to overview

Multi-level modeling of molecular circuitry to facilitate precision medicine of complex disease

Applicant Gruber Andreas
Number 178591
Funding scheme Early Postdoc.Mobility
Research institution Big Data Institute University of Oxford
Institution of higher education Institution abroad - IACH
Main discipline Cancer
Start/End 01.10.2018 - 31.03.2020
Show all

All Disciplines (2)

Discipline
Cancer
Clinical Cancer Research

Keywords (11)

Personalized medicine; Precision medicine; Cancer; Complex disease; Clinical decision making; Machine learning; Artificial intelligence; Genomics; Transcriptomics; Deep learning; Drug target identification

Lay Summary (German)

Lead
In den letzten beiden Jahrzehnten wurde die Technologie der Hochdurchsatz-Sequenzierung stark verbessert. Der technologische Fortschritt in diesem Gebiet ermöglicht es, dass heute die Sequenz von menschlichen Genomen mit relativ geringem Aufwand gelesen und die Aktivität der Gene gemessen werden kann. Nachdem die Technologie noch relativ jung ist, kann aktuell nur ein Bruchteil der enthaltenen Informationen interpretiert und für Forschung, Diagnose und klinische Entscheidungsprozesse genutzt werden.
Lay summary
Eine große Anzahl an unterschiedlichen Molekülen, deren Bauplan im Genom festgeschrieben ist, bilden die Bauteile von hoch komplexen, molekularen ‘Schaltkreisen’, welche der Zelle ein gerichtetes Reagieren auf intra- und extrazelluläre Reize ermöglichen. Bei ‘komplexen’ Krankheiten, wie zum Beispiel Krebs, kommt es zu einer Vielzahl an genomischen Veränderungen (Mutationen), welche in Ausprägung und Kombination in der Regel für jeden Patienten/jede Patientin einzigartig sind. Die sogenannte personalisierte Medizin, auch Präzisionsmedizin genannt, versucht die genetischen Besonderheiten von Patienten/Patientinnen zu berücksichtigen. Ziel des Projektes ist die Entwicklung eines auf Maschinellem Lernen basierenden Analyseverfahrens, welches dazu in der Lage ist, die großen Datenmengen zu nutzen, um (i) deregulierte ‘Schaltkreise’ zu identifizieren, (ii) patienten-spezifische Veränderungen in Relation zu Daten anderer Patienten/Patientinnen zu setzten und (iii) die Form, die Ausprägung und/oder den Grad der Krankheit entsprechend zu klassifizieren.

Das Projekt wird sowohl die molekularbiologische Erforschung von Krankheiten, als auch die personalisierte Medizin vorantreiben, indem es (i) neue Einblicke in die molekularen Mechanismen bietet, die bestimmten Krankheiten zugrunde liegen und (ii) Patienten/Patientinnen anhand molekularer Merkmale klassifiziert.
Direct link to Lay Summary Last update: 29.08.2018

Responsible applicant and co-applicants

Name Institute

Publications

Publication
Reply to ‘A different perspective on alternative cleavage and polyadenylation’
Gruber Andreas J., Zavolan Mihaela (2020), Reply to ‘A different perspective on alternative cleavage and polyadenylation’, in Nature Reviews Genetics, 21(1), 63-64.
PolyASite 2.0: a consolidated atlas of polyadenylation sites from 3′ end sequencing
Herrmann Christina J, Schmidt Ralf, Kanitz Alexander, Artimo Panu, Gruber Andreas J, Zavolan Mihaela (2019), PolyASite 2.0: a consolidated atlas of polyadenylation sites from 3′ end sequencing, in Nucleic Acids Research, 48(D1):D17.
Alternative cleavage and polyadenylation in health and disease
Gruber Andreas J., Zavolan Mihaela (2019), Alternative cleavage and polyadenylation in health and disease, in Nature Reviews Genetics, 20(10), 599-614.
Terminal exon characterization with TECtool reveals an abundance of cell-specific isoforms
Gruber Andreas J., Gypas Foivos, Riba Andrea, Schmidt Ralf, Zavolan Mihaela (2018), Terminal exon characterization with TECtool reveals an abundance of cell-specific isoforms, in Nature Methods, 15(10), 832-836.

Collaboration

Group / person Country
Types of collaboration
Prof. Mihaela Zavolan, Biozentrum, University of Basel Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof. Ludmil Alexandrov, University of California, San Diego United States of America (North America)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof. Ian Tomlinson, MRC Institute of Genetics & Molecular Medicine, Cancer Research UK Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Dr. David Church, Wellcome Centre for Human Genetics, University of Oxford Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof. Richard Houlston, Institute of Cancer Research Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof. Erik van Nimwegen, Biozentrum, University of Basel Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
Invited Talk at the Center for Bioinformatics Individual talk Multi Level Modeling of Molecular Circuitry in Health and Disease 03.03.2020 Saarland University, Germany Gruber Andreas;
Invited Talk at the Department of Mathematics and Computer Science Individual talk AI-aided molecular multi-level modeling to advance precision medicine 09.12.2019 Department of Mathematics and Computer Science, University of Southern Denmark, Denmark Gruber Andreas;
EMBO Workshop on Precision Health Poster The molecular landscape of endometrial carcinoma 13.11.2019 EMBL, Heidelberg, Germany Gruber Andreas;
Invited Talk at the Institute of Molecular Cancer Research Individual talk Molecular multi-level modeling to advance precision oncology 05.11.2019 Faculties of Science and Medicine, University of Zurich, Switzerland Gruber Andreas;
Genomics England Research Conference Poster The molecular landscape of endometrial carcinoma 04.11.2019 London, Great Britain and Northern Ireland Gruber Andreas;
Invited Talk at the Symposium on Artificial Intelligence in Medicine Individual talk Molecular multi-level modeling to advance precision oncology 17.07.2019 Medical Research Centre of the University Hospital Essen, Germany Gruber Andreas;


Self-organised

Title Date Place
Oxford BDI Network Meeting 13.02.2020 University of Oxford, Great Britain and Northern Ireland
International Workshop on Data Science Challenges 30.01.2020 University of Oxford, Great Britain and Northern Ireland
BDI Socials 30.01.2019 University of Oxford, Great Britain and Northern Ireland

Knowledge transfer events



Self-organised

Title Date Place
Social Science Fridays 08.02.2019 Oxford, Great Britain and Northern Ireland

Use-inspired outputs

Software

Name Year
PolyAsite 2.0 Database 2020
TECtool 2018


Abstract

Within the past decade next generation sequencing technologies have enabled a systems perspective on gene expression and its regulation at various co- and post-transcriptional levels. Systems biology studies have provided fundamental insights into cell biology and drastically improved our understanding of the complexity of human disease, including cancer. Between-patient tumor heterogeneity indicates that therapeutic successes can be improved through molecular characterization, personalized forecasts of a patient course of disease and personally tailored therapeutic decisions.Within the project proposed here we aim to implement a system, termed MoCi-PMS system (for Molecular Circuitry-based Precision Medicine Support system), which will elucidate molecular circuitries that are deregulated within patients. Making use of a multitude of sophisticated computational approaches developed over recent years MoCi-PMS will provide integral insights into complex disease and support precision medicine research as well as clinical decision making. In particular, for a given patient the MoCi-PMS system will (i) identify deregulated molecular circuitries at multiple regulatory levels, (ii) compare patient-specific molecular alterations to the ones found in other patients having similar disease patterns and (iii) estimate the probability of the patient to have the assumed disease and a specific pathological stage, respectively, making use of a computational approach that integrates predictive features inferred from multiple types of data, including cancer stage estimates, whole-genome sequencing and RNA sequencing.MoCi-PMS feeds directly into current efforts to advance precision medicine by providing fine-grained, multi-level analyses of complex disease thereby facilitating clinical decision making, but also drug target discovery. By identifying features that are common and predictive for the pathological stage of patients having similar disease patterns, the system will provide novel insights into the molecular mechanisms that are characteristic to specific pathologies, thereby revealing actionable signatures that will feed into research of patient-tailored therapies.
-