Senior Data Scientist at Bio-Prodict

I am currently employed at Bio-Prodict as a Senior Data Scientist, where I use state-of-the-art machine learning techniques to develop novel solutions for bioinformatics problems.

I am primarily involved in the production of the Helix product, leading a team that builds on the results of previous research to predict pathogenicity for different protein variants.

Key Responsibilities

• Developing AI-based solutions in Bioinformatics, including exploratory research, system design, and hardware provisioning.
• Building and adapting state-of-the-art machine learning models.
• Deploying solutions to on-premise hardware or cloud environments.
• Designing and implementing interfaces for solutions.
• Leading and mentoring students and junior team members.
• Conducting direct client communications, including sales talks and technical support.
• Representing Bio-Prodict at international scientific conferences.
• Maintaining and installing on-premises Helix servers.
• Managing Helix cloud capabilities.

Skills

Technical Expertise

• Expert in Python (PyTorch, Keras, TensorFlow, scikit-learn, numpy, scipy, matplotlib)
• Experience with Rust, PHP, Java, SQL, R, HTML/CSS/JS
• Git CI/CD pipelines
• Google Cloud Platform (Kubernetes, Cloud Run, Cloud SQL)
• Relational and Document-based databases (PostgreSQL, MySQL, MongoDB)
• ElasticSearch

Publications

1. Heijl S, Vroling B, van den Bergh T, Joosten HJ. (2020). Mind the gap: preventing circularity in missense variant prediction. bioRxiv, 2020.05.06.080424. Cited by: 4

2. Vroling B, Heijl S. (2021). White paper: the helix pathogenicity prediction platform. arXiv preprint arXiv:2104.01033. Cited by: 7

3. Heijl S, Boot J, Bergh T, Vroling B, Joosten HJ, Brier B. (2022). Helix engineering: Combining the power of 3DM with AI to disrupt protein engineering.

4. Boonen RACM, Wiegant WW, Celosse N, Vroling B, Heijl S, Kote-Jarai Z, et al. (2022). Functional Analysis Identifies Damaging *CHEK2* Missense Variants Associated with Increased Cancer Risk. Cancer Research, 82(4), 615-631. Cited by: 40

5. Dorling L, Carvalho S, Allen J, Parsons MT, Fortuno C, González-Neira A, ..., Heijl S, et al. (2022). Breast cancer risks associated with missense variants in breast cancer susceptibility genes. Genome Medicine, 14(1), 51. Cited by: 25

6. Ramakrishnan G, Baakman C, Heijl S, Vroling B, van Horck R, Hiraki J, et al. (2023). Understanding structure-guided variant effect predictions using 3D convolutional neural networks. Frontiers in Molecular Biosciences, 10, 1204157. Cited by: 7

7. Joosten HJ, Vroling B, Heijl S, van den Bergh T. (2023). Transforming protein engineering: advanced integration of deep learning and 3DM technology for superior protein function predictions.

8. Vroling B, van den Bergh T, Alders M, Heijl S, Tanck M, Deprez RLD, et al. Inherited arrhythmia syndromes, how to identify pathogenic mutations?