Etienne Feyrer

Bioinformatics Graduate, University of Tübingen
Background in applied Machine Learning, Contrastive learning and Vector similarity search

Vancouver, Canada

About Me

I am building tools for computational biology and clinical genomics. In particular, I am experienced in develping faster and more efficient Machine learning methods, drawing from my knolwedge in vector search, database architectures and parallel computing.

For my undergraduate research thesis at the HassounLab, I developed a machine-learning-based analogue compound search system, leveraging efficient vector search to accelerate molecular similarity retrieval. At the Institute of Medical Genetics and Applied Genomics, I designed and implemented a full-stack application supporting a next-generation sequencing (NGS) pipeline. I also developed software for the curation and management of a large-scale online database in an e-commerce setting.

Projects

I develop open-source tools that tackle challenges in computational biology, NGS analysis, and health sciences, combining algorithmic methods, machine-learning techniques, and solid software engineering.

SpectraSnap

End-to-end metabolomics annotation pipeline built on the JESTR framework. It uses FAISS for fast, scalable retrieval and supports preprocessing, embedding, and annotation of untargeted MS/MS data. The project has been validated on a real-world breast cancer dataset, demonstrating strong performance in practical metabolomics workflows.

Skin Lesion Classifier

A machine-learning application that analyzes smartphone images of skin lesions together with patient data to classify them as benign or malignant, using Multimodal Ensemble Learning with DenseNet

Clinical prediction tool

A clinical variant analysis and decision-support tool for the Institute of Medical Genetics and Applied Genomics, that annotates VCF variants using Ensembl VEP, caches results in MariaDB, and scores pathogenicity with a Random Forest meta-model. Tool integrates GUI for clinicians or bioinformaticians to load VCFs, call the annotation API, filter/prioritize variants, and export results.

Work and Achievements

I spent over a year as an Automation Engineer for a major e-commerce platform, building scalable workflows and improving operational efficiency. I earned a competitive room-and-board scholarship to study at Tufts University. I completed my B.Sc. in Bioinformatics at the University of Tübingen, a designated German University of Excellence with leading research clusters in life sciences and computational biology.

engineering

Automation Engineer

I built a new data-processing system for a high-traffic e-commerce platform, standardizing multi-vendor data and cutting processing time from 6 hours to 30 minutes through asynchronous task orchestration.

workspace_premium

Tufts scholarship

Awarded a competitive study-abroad scholarship at Tufts University, completing advanced coursework in Computational Systems Biology and Machine Learning with a 3.92 GPA.

school

B. Sc. Bioinformatics

Graduated in Bioinformatics from the University of Tübingen, a German University of Excellence, with advanced training in computation theory algorithmic foundations and Bioinformatics.

Skills

I specialize in deep learning, representation learning, contrastive learning, vector similarity search, genomic variant analysis, and predictive modeling. I have strong software engineering skills with hands-on experience in Python, PyTorch, REST APIs, Docker, SQL databases, Linux, and scalable data-processing pipelines.