PhD Position in Deep Learning Modeling of Cancer Karolinska Institutet in Sweden

Karolinska Institutet, one of the world's leading medical universities, is offering a PhD position in deep learning modeling of cancer. The position is based in Avlant Nilsson's research group at Karolinska Institutet and SciLifeLab, focusing on the development of mechanistic AI models of cancer cells to advance precision medicine. The group integrates large-scale multi-omics datasets (such as metabolomics, transcriptomics, and proteomics) with molecular interaction networks to build interpretable deep learning models that describe cancer mechanisms at the systems level. The overarching goal is to create a unified model explaining how molecular processes like signaling, gene regulation, and metabolism interact to determine cellular behavior, aiming to identify biomarkers, new drug targets, and resistance mechanisms, and ultimately enable computer-aided design of precision treatments.

The project is part of the SciLifeLab and DDLS (Data-Driven Life Science) program, providing access to a highly collaborative and interdisciplinary research environment and advanced computational infrastructure, including the Berzelius supercomputer. The main supervisor is Assistant Professor Avlant Nilsson, with Professor Randall S. Johnson as co-supervisor at the Department of Cell and Molecular Biology.

The doctoral project addresses the dynamic nature of cell metabolism in cancer, a hallmark of tumor development. Traditionally, metabolism, signaling, and gene regulation have been studied with separate modeling frameworks, limiting our understanding of their interplay and the prediction of metabolic effects of drugs targeting cellular signaling. In this PhD project, you will develop a next-generation AI model of the cancer cell by integrating metabolism, signaling, and gene regulation into a unified predictive framework. The work involves developing biologically constrained deep learning models based on a novel architecture with recurrent neural networks (RNNs) to describe how molecular states evolve over time and give rise to cellular phenotypes.

The project consists of three main parts: (1) Model development—integrating modules for cellular subsystems to predict cellular states; (2) Target identification in silico—predicting metabolic vulnerabilities and prioritizing candidates for experimental validation; (3) Biomarker identification—identifying molecular states linked to drug response in various cancer contexts. You will work closely with the computational research group and collaborators, contributing to the design of experiments based on model predictions.

Karolinska Institutet offers a creative, international, and inspiring environment filled with expertise and curiosity. As a PhD student, you will have an individual research project, a skilled supervisor, a wide range of elective courses, and the opportunity to work in a successful research group. KI collaborates with leading universities worldwide, offering opportunities for international exchanges. As an employed doctoral student, you receive a contractual salary, free access to modern fitness facilities, and compensation for medical care.

Eligibility requirements include a master's degree or equivalent, or at least 240 ECTS credits with 60 at the advanced level, or equivalent knowledge. English proficiency equivalent to English B/English 6 is required. Strong programming skills (e.g., Python), experience with version control (e.g., Git), ability to work independently and in teams, and good communication skills in English are necessary. Merits include background in physics, computational biology, machine learning, applied mathematics, experience with deep learning frameworks (e.g., PyTorch, TensorFlow), knowledge in cancer and cell biology, experience with high-dimensional or dynamic systems, and analysis of biological or multi-omics data. Prior biology experience is not required; applicants from quantitative disciplines are encouraged. The group values diversity and inclusion, fostering an environment where all individuals are empowered to contribute their unique perspectives and skills toward the shared goal of advancing precision medicine in cancer.

The doctoral position is a full-time employment for up to 4 years. The application deadline is May 27, 2026. Apply via the Varbi recruitment tool, submitting your application in English or Swedish, including a personal letter, CV, any theses and publications, documents verifying required skills and qualifications, and proof of eligibility. Follow instructions on the doctoral education eligibility web pages. All applicants will be informed when the recruitment is completed.