Tim Huygelen

About me

Hi, I'm Tim, a recent UCL Biochemistry MSci graduate with a strong focus on bioinformatics and machine learning. My work combines molecular biology with data science, such as building machine learning models and automating processes to solve complex biological problems with Bayesian optimisation. I've gained hands-on experience with Python, R and SQL, and have worked on exciting projects ranging from developing tools for DNA sequence generation to optimising cell tracking algorithms.

In addition to my academic work, I'm passionate about creative projects - I've organised electronic music events and started my own digital art series. I love working at the intersection of science and creativity.

What i'm doing

Bioinformatics Research

Using advanced computational tools to analyze biological data, from large-scale DNA sequences to protein structures.
Machine Learning

Developing and optimising machine learning models for biological applications.
Data Analysis & Visualisation

Performing in-depth data analysis and creating clear, impactful visualizations to communicate complex biological insights effectively.
Digital Art

I make abstract art in photoshop, I think it looks pretty neat!

Testimonials

Dr. Alan Lowe

Dr. Alan Lowe
Anna Maria Papaoikonomou

Tim was such a joy to work with at Women's Wrongs, our feminist zine society at UCL! His creativity and outgoing personality really brought the group to life, and he was always open to new ideas. He was incredibly open-minded and committed to making everyone feel included, always showing empathy and taking the time to understand people, especially those from different cultural backgrounds. The space felt super welcoming thanks to his communication skills, and he made running the zine look effortless. We all had a great time working together!
César García Herrera

César García Herrera

Organisations

CV

Education

University College London (UCL)
2020 — 2024

Biochemistry MSci (Molecular Biology)

Final Year Mark: 75.83%, Upper Second-Class Honours

Notable and Elective Modules Taken:

Computational and Systems Biology (SQL, Perl), Introduction to Coding for Bioscience Research (Python), Specialist Research Project in Metagenomics (R), Rethinking Capitalism, Biomolecular Structure and Function, Economic Geography II, Immunology, Advanced Molecular Biology of Protein Regulatory Networks
Koninklijk Atheneum Tervuren, Belgium
2018 — 2020
General Secondary Education (Maths and Sciences)
The German School London, UK
2017 — 2018
Gymnasium (English and German taught at native level)

Research Experience

4th Year Dissertation
2023 — 2024
University College London – Supervisor: Dr. Alan Lowe (October 2023 – May 2024)

- I used the Optuna Python library to construct an automated parameter tuning framework for a deep-learning based cell tracking software “btrack” to improve overall accuracy and generalisability for any cell tracking dataset.

- Demonstrated that my automated parameter tuning implementation significantly reduces reliance on manual input, while vastly improving its reliability, generalisability, and accuracy.

- Investigated various optimisation strategies, including Tree-Structured Parzen Estimators, and two evolutionary parameter tuning algorithms to find one most suitable to our problem.

- Successfully implemented a pruning strategy for the hyperparameter optimisation and implemented a tool allowing for quick visualisation of constructed lineage trees.
Diffusion Model Development for Promoter Sequence Generation
2023
Imperial College London – Prof. Guy-Bart Stan Group (July 2023 – September 2023)

- Awarded a grant by the Biochemical Society for a summer research initiative, where I:

- Developed EPD-GenDNA, the first comprehensive multi-species dataset tailored for DNA sequence generation. This dataset contains 160,000 unique DNA sequences from 15 different species, spans 2,713 cell types, and is underpinned by 30 million wet lab experiments.

- Designed and implemented the UNet architecture within DiscDiff, an innovative tool employing latent diffusion for the generation of DNA promoter sequences.

- Contributed as third author to the publication of "Latent Diffusion Model for DNA Sequence Generation". This paper, which has received 3 citations, was presented at NeurIPS, an internationally renowned AI conference.
3rd Year Dissertation
2022 — 2023
University College London – Supervisor: Dr. Alan Lowe (October 2022 – May 2023)

- Literature review on the revolution in deep-learning based protein design software after the release of AlphaFold 2

- My research included: structure prediction software-based backbone hallucination, diffusion models (e.g. RF diffusion), sequence generation software (e.g. ProteinMPNN), and physics-based tools.

- Examined three case studies of groundbreaking research of successfully designed proteins to better understand applied use of traditional and new deep learning-based protein design tools.

- Finally, I proposed a community-wide protein design competition to accelerate protein design research.
Research Project in genetics on SARS-CoV-2 spike protein
2021
UCL – van Dorp Group (July 2021 – December 2021)

- I collaborated with Prof. François Balloux and Dr. Lucy van Dorp to investigate the potential correlation between the transmission rates of SARS-CoV-2 variants and the affinity of their spike proteins for the ACE2 receptor. To accomplish this, I created a data analysis tool in R, which processes extensive multiple sequence alignments (MSAs) obtained from GISAID and generates the distribution of predicted binding affinities for each variant. I found evidence suggesting a significant correlation between binding affinity and transmission rates.