PhD Position in Computational Methods with link to Tumorigenesis
Catholic University of Leuven Department of Oncology
Belgium

Project

The prevailing genetic theory of cancerogenesis postulates that malignant transformations is driven by an accumulation of pathogenic mutations that gradually increase the fitness and proliferative potential of cells that give rise to cancer [1]. However, recent studies have demonstrated that common genetic drivers are often mutated in healthy tissues [2], and non-genetic factors, such as epigenetic heterogeneity, together with the cues from tissue microenvironment, might be playing a crucial role in cancer initiation [3]. In order to resolve this ambiguity for the benefit of cancer prevention,early detection and treatment, one would have to recover the earliest gene regulatory landscapes of cancer cells-of-origin and separate the oncogenically induced alterations from those that emerged due to intrinsic heterogeneity and the impact of microenvironment. The project will aim at developing computational methods to (i) deconvolute tumor,immune and stromal compartments in bulk, single-cell and spatial omics data of solid cancers, (ii) reconstruct the transcriptional and gene regulatory states of pre-malignant and early neoplastic cells, and (iii) disentangle the role and contributions of the germline genetic backround, somatic mutations and non-genetic factors to neoplastic transformation. Based on our previous work with latent factor analysis in epigenomic data [4,5], we will apply classical(e.g. NMF) and cutting-edge (e.g. deep CNNs, autoencoders) machine learning approaches to jointly model data of various modalities. We will combine bulk omicsprofiles from large patient cohorts with time-resolved single-cell and spatialomics maps to bring together statistical power and high cellular resolution in selected samples, and integrate gene expression maps with epigenomic patterns(e.g. chromatin accessibility and DNA methylation) to pinpoint the gene regulatory mechanisms driving tumorigenesis and develop novel powerful biomarkers. Through own and our collaboration partners’ entity-focused projects, we will validate our methodology in several cancer types, including colorectal, pancreatic, prostateand lung cancers, as well as rare sarcoma entities [6].

References:

1.       Vogelstein et al. Cancer genes and the pathways they control. Nature Medicine, 2004.

2.       Lawson etal. Extensive heterogeneity in somatic mutation and selection in the human bladder. Science, 2020.

3.       Alonso-Curbelo et al. A gene-environment-induced epigenetic program initiates tumorigenesis. Nature, 2021.

4.       Lutsik et al. MeDeCom: discovery and quantification of latent components of heterogeneous methylomes. Genome Biology, 2017.

5.       Sherer et al. Reference-free deconvolution, visualization and interpretation of complexDNA methylation data using DecompPipeline, MeDeCom and FactorViz. Nature Protocols, 2020.

6.       Lutsik et al. Globally altered epigenetic landscape and delayed osteogenic differentiation in H3.3-G34W-mutant giant cell tumor of bone. Nature Communications, 2020.

Profile

We are searching for a highly motivated candidate with specialized training (Masters degree or equivalent) in computational biology, bioinformatics, computer science or related disciplines. We expect:
- solid knowledge of state-of-the-art machine learning and data science methodology,combined with a genuine interest to leverage it for solving important problems in cancer research (previous experience single-cell and/or spatial omics data is a major advantage);
- excellent programming skills in one of the scripting (Python, R) and/or core development (C/C++/Java) languages, as well as experience with software packaging and deployment;
- fluency with Linux-based computational infrastructure and familiarity with majorHPC enviornments (cluster schedulers, cloud computing etc);
- willingness to gain deep biological understanding of the studied cancer entities;
- well-structured working style and appreciation for reproducible research practices in computational biology, e.g. proper data management and scientific workflows;
- goal-oriented team player attitude and ability to spearhead interactions with collaboration partners;
- good communication skills and command of English language, both spoken and written.

Offer

We offer a full time PhD position for 48 months,given positive evaluation by the doctoral committee at the end of the first year. A successful PhD candidate will be immersed into a vibrant and highly stimulating scientific environment within the Department of Oncology, VIB-KULeuven Center for Cancer Biology and beyond, with ample possibilities to interact with researchers from other groups. For their project the candidate will obtain access to cutting-edge datasets and computational infrastructure. The offered project is highly collaborative and will involve participation in the work of global consortia, such as Pan-Prostate Cancer Group, regular exchange visits with collaboration partners across Europe, presentations at major conferences and meetings. Beyond that, you will have access to versatile options for personal growth and development, advanced qualification courses and mentoring.

Interested?

For more information please contact Prof. dr. Pavlo Lutsik, mail: pavlo.lutsik@kuleuven.be 

If you apply for this position please say you saw it on Medicinoxy