Valentine Svensson
Principal Computational Biology Scientist
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolValentine Svensson is a Principal Computational Biology Scientist in San Francisco with 14 years of experience applying statistical and machine learning methods to cell and single-cell genomics. He blends a strong quantitative foundation—PhD in Molecular Biology from Cambridge and an M.Sc. in Mathematics—with hands-on software engineering, contributing to open-source bioinformatics projects like scvi-tools (LDVAE for single-cell RNA-seq) and bcbio-nextgen. His work spans production pipelines, configuration and DevOps improvements, and probabilistic deep learning models, reflecting both reproducible engineering and biological insight. At startups and research labs (Altos Labs, Vesalius, Caltech, EMBL-EBI) he’s led analysis of spatial and single-cell assays and optimized high-throughput backends. Notably, he writes robust backend integrations (e.g., RabbitMQ/CouchDB logging) and emphasizes practical, well-tested tools that make advanced genomics methods accessible to teams.
14 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Molecular Biology, Doctor of Philosophy - PhD Molecular Biology at University of CambridgeMaster of Science (M.Sc.) Mathematics, Master of Science (M.Sc.) Mathematics at Lund University
Swedish, English
Github Skills (28)
pytorch10
kombu10
python10
pandas10
seq10
configuration-management10
machine-learning10
pipelining10
message-queue10
sc10
rabbitmq10
deep-learning10
fileio10
pipe10
pipeline10
Programming languages (9)
Github contributions (5)
scverse/scvi-tools
May 2019 - Jun 2020
Deep probabilistic analysis of single-cell and spatial omics data
Role in this project:
Data Scientist
Data ScientistContributions:23 commits, 3 PRs, 48 comments in 1 year 1 month
Contributions summary:Valentine's primary contribution involves the development and implementation of a linearly decoded variational autoencoder (LDVAE) model, a deep learning approach for analyzing single-cell RNA-seq data. They refactored existing code, created a new class for the LDVAE, and added a notebook demonstrating its usage and interpretation of gene relationships. The user also added functionality to extract linear loadings and conducted testing to validate the model's performance across different batch scenarios using the pbmc3k dataset.
hierarchicaldeep-generative-modelmixture-of-expertssingle-cell-rna-seqsingle-cell
chapmanb/bcbb
Sep 2011 - Mar 2012
Incubator for useful bioinformatics code, primarily in Python and R
Role in this project:Back-end Developer
Back-end DeveloperContributions:7 commits in 6 months
Contributions summary:Valentine contributed significantly to the configuration and functionality of the bioinformatics pipeline. They implemented a module for loading configuration from YAML files, expanding environment variables, and applied it across several scripts. Further work generalized the configuration loading, incorporating path expansion and handling for various settings within the configuration. Additionally, the user made enhancements to existing code, such as adapting the output of the metrics report for different decimal formats.
pythonpython-bindingsgenomicsr-and-pythonbioinformatics
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.