› georgy klochkov · computational rna biology · ulm, germany

We build reproducible pipelines for RNA-folding.

Twenty years of integration and infrastructure engineering in banking and industry, now applied to computational RNA biology. My focus is making machine-learning results on RNA structure checkable by other people.

How much accuracy survives homology-aware splits?
• in progress  [status: draft / preprint live]

RNA secondary-structure models are usually evaluated on random splits. This project asks how much of their accuracy remains when training and test data are separated by RNA family, and provides a minimal pipeline so anyone can rerun the answer.

cobalt is a small tool for sequence handling and quality control. It reads input sequences, checks them, and passes clean, consistent records downstream. In the project above, it is the input layer of the pipeline.

[repository link]

I trained in computing systems engineering at Bauman Moscow State Technical University. Since then I have worked on enterprise integration, data flows and software architecture across banking, industry and consulting, including team lead and architect roles. I am now moving my experience into computational biology, where reproducibility is still an open problem as much as application of new approaches. I am based in Ulm and open to hybrid or remote work.
Gerne auch auf Deutsch.

Earlier: software and integration engineering in banking, automotive and embedded systems, 2007–2026.

# skills
Python, R, workflow managers, containers, ViennaRNA, Git and CI
Java, C++, Rust, SQL, AWS

Interested in research positions in computational RNA biology. Based in Ulm or Tübingen area, open to hybrid or remote.