We will continue to advocate with public bioinformatics institutions to correct the annotation of selenoproteins, and eventually other cases of translational recoding, in public databases

To fix this issue, we present an upgraded version of our pipeline Selenoprofiles to automatically and accurately annotate selenoproteins at scale, matching manual curation quality. This may help to uncover overlooked biology tied to redox homeostasis and protein function.

We show that the great majority of vertebrate selenoprotein genes are misannotated in major public databases, meaning that they appear often missing or incorrect. Since annotations are a foundational resource upon which much of bioinformatics is built, these flaws impact many downstream analysis.

Selenoproteins are special proteins incorporating the amino acid selenocysteine, which play many essential biological roles, most notably for antioxidant defense. Selenocysteine is encoded by the UGA stop codon via translational recoding, making selenoproteins notoriously hard to annotate.

We welcome feedback and contributions! Use GitHub to report issues or request features, and to follow our Continuous-Integration-compliant development: github.com/pyranges/pyr...
Wannabe contributors can find guidelines here: pyranges1.readthedocs.io/en/latest/de...

Pyranges works best as python library, but if you are a die-hard bash enthusiast that despises Python for some reason, you may access its functionalities from the command line, thanks to pyranger, a Pyranges wrapper based on Google Python Fire library.
pyranges1.readthedocs.io/en/latest/py...

Developed for max speed, Pyranges is also suitable for interactive data exploration, supported by convenient data access functions and intuitive Jupyter-compatible representations, including an early prototype of the companion graphics library Pyrangeyes 🐍👀(pyrangeyes.readthedocs.io).

Pyranges is ready for AI-assisted coding. It provides utilities to turn chatbots (e.g. ChatGPT, Gemini) into specialized coding assistants that can build readable Pyranges-based workflows from simple user requests. Guide: pyranges1.readthedocs.io/en/latest/tu...
Examples of AI-generated workflows:

Our library is designed to be user-friendly. The PyRanges object (representing a table of genomic interval-based data) is a subclass of pandas.DataFrame, allowing users to seamlessly leverage the familiar pandas API while combining it with Pyranges-specific methods. See how in our linked tutorial!

With big data, Pyranges is ~5× faster than BEDTools and ~3× faster than GenomicRanges, largely due to its Rust-based core engine, ruranges. In addition, Pyranges uses about half the memory required by these alternative tools. See the preprint for details.

This library allows you to read lots of diverse omics data (e.g. gene annotations, mapped NGS reads, chromosome and protein domains, in many file formats) and readily build complex pipelines by combining efficient methods for overlapping, counting, slicing intervals, and much more. Cheatsheet below.

Learn more on the wiki: 2025.igem.wiki/barcelona-ub/
Video presentation (15 min): video.igem.org/w/c6034b4e-8...

Skippit fuses a SECIS element (readthrough driver) with a theophylline-binding aptamer, yielding the first readthrough-based riboswitch.
By changing ligand concentration, you can flip between producing the canonical protein isoform or the readthrough isoform:
video.igem.org/w/rDM2PWQi8Y...

The project Skippit uses stop codon readthrough as a synthetic biology control knob. In readthrough, the ribosome occasionally ignores a stop codon and keeps translating — extending the protein instead of terminating it:
video.igem.org/w/9U3qG3tzMA...