Project title
Development of the ADvanced mAchine learning Precision Treatment Strategy (ADAPTS) platform for patients with Diffuse Midline Glioma (DMG).
Collaborators and funding
Collaborators:
-
Fa Valdes-Mora, Cancer Epigenetic Biology and Therapeutics group, Children’s Cancer Institute, Australia
-
Dieter Henrik Heiland, The MILO laboratory, University of Freiburg, Germany
-
Nada Jabado, Nada Jabado Laboratory, McGill institute, Canada
-
Sebastian Waszak, Waszak Lab, École Polytechnique Fédérale de Lausanne, Switzerland
-
Sabine Mueller, University of California San Fransisco, United States
-
Nick Vitanza, Vitanza Lab, Seattle Children’s Research Institute, United States
-
Rob Salomon, University of New South Wales, Australia
-
Mark Cowley, Children’s Cancer Institute, Australia
-
Santosh Valvi, The University of Western Australia, Australia
Bioinformatics support:
-
Digital Technology Solutions, Resources Division, University of Newcastle, Australia
Funding partners: Little Legs Foundation and RUN DIPG.
Contact(s)
-
Project lead: Matthew D. Dun, University of Newcastle, Australia, matt.dun@newcastle.edu.au
-
Project co-lead: Fatima Valdes Mora, University of New South Wales, Sydney, Australia, fvaldesmora@ccia.org.au
-
Computational systems lead: Clara Savary, University of Newcastle, Australia, clara.savary@newcastle.edu.au
-
Bioinformatics lead: Tuan Vo, University of Newcastle, Australia, tuan.vo@newcastle.edu.au
Project description and aims
Diffuse midline glioma (DMG) is a uniformly fatal brain tumour, representing the most common cause of cancer-related death in children. Despite extensive research efforts, radiotherapy remains the sole treatment option and patients still tragically succumb within a year of diagnosis. It remains clear that in the clinical setting, our current armamentarium against DMG falls desperately short of acceptable, highlighting the urgent imperative for innovative research and therapeutic strategies.
This project aims to develop the ADvanced mAchine learning Precision Treatment Strategy (ADAPTS) platform, which will enable the analysis of time-series multiomics data from core biopsies, cerebrospinal fluid, and peripheral blood samples collected from both on-treatment patients and DMG mouse models. This framework will enable to model tumour adaptation to treatments and identify the next targetable vulnerability. To achieve this, machine learning models are trained on on-treatment matched tumour and blood samples to help predicting the relevant timeframe for administering the subsequent treatment based on blood profiling. The objective of this project aligns with the era of personalised medicine, as we aim to adjust therapeutic regimens throughout the treatment journey of patients using non-invasive techniques to enhance their survival and quality of life.
How is ABLeS supporting this work?
This work is supported through the Production Bioinformatics scheme provided by ABLeS. The support includes 60 TB long term storage and 50 KSUs per quarter.
Expected outputs enabled by participation in ABLeS
A public ADAPTS platform, featuring computational workflows developed on GitHub, and freely accessible datasets available for researchers and clinicians to analyse the temporal adaptation of therapeutics used in DMG treatment. The machine learning models trained on these data are anticipated to inform the clinical management of DMG patients by suggesting a temporal and sequential treatment plan tailored to each patient’s tumour biology in real-time using non-invasive blood profiling. Our goal is to publish our findings in high-impact journals such as Nature and Cancer Cell.
These details have been provided by project members at project initiation. For more information on the project, please consult the contact(s) or project links above.