iiinova — where your digital self keeps evolving

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

GOOSE has no fancy hardware requirements and runs across different platforms. Documentation is here goose.readthedocs.io and colab notebook here colab.research.google.com/drive/1U9B-T...

goose.readthedocs.io

Welcome to goose’s documentation! — goose documentation

000

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Overall, GOOSE makes it easy for anyone to start designing libraries of sequences to investigate their favourite IDR or to test an emerging hypothesis. We are also – of course – happy to offer guidance and help in library design.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Finally, we designed a library of >2000 sequence variants to systematically explore the molecular determinants of how disordered proteins convey desiccation tolerance in yeast. See thread from @karaehunt.bsky.social for more on this (bsky.app/profile/kara...).

karaehunt.bksy.social

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

We then design a 2-component in-cell recruitment system in which one IDR forms cellular assemblies, and another is targeted to (or excluded from) those assemblies, all using de novo IDRs designed from scratch (200- and 400-residue proteins).

110

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

We also designed additional sequences with designated ensemble dimensions and explored the relationship between sequence dimensions and in-cell crowding, see thread from @trevorbrandt.bsky.social for more on this (bsky.app/profile/trev...).

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

First, we designed a set of 32 de novo IDRs to explore sequence influences IDR ensemble dimensions in cells, finding good agreement with prior in vitro and in silico conclusions, with the exception of sequences with lots of positively charged residues.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Here, we deploy GOOSE across a range of vignettes to both demonstrate its versatility and explore some key questions of interest to us.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Further, for built-in design objectives, GOOSE enables large (10,000+) sequence libraries to be generated in minutes

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Sequences can be designed to vary overall properties, IDR dimensions, intermolecular interactions, or really any arbitrary function via our general SequenceOptimizer.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

This is where GOOSE comes in. GOOSE is a general computational design framework that allows you to design IDRs in many different ways:

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

By designing sequence libraries that are large (10s, 100s, 1000s, 10,000s of sequences) and testing these all at once, we can test multiple interpretable hypotheses in parallel.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

We (and others!) propose that for IDRs, rational sequence design offers a general route towards sequence-to-function mapping. By designing libraries of proteins that systematically vary different aspects of the sequence, we can explore how those aspects influence the output of a readout or assay.

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Disordered regions are often (not always!) critical for cellular function, but practically, how do we explore this link between sequence and function without a reference structure to guide mutations?

100

Ryan Emenecker @remenecker.bsky.social · 27/02/2026

Excited to share our substantially updated manuscript, "Rational design of disordered proteins for systematic sequence-to-function investigation"! www.biorxiv.org/content/10.1...

biorxiv.org

172

Ryan Emenecker @remenecker.bsky.social · 16/02/2025

@bornanovak.bsky.social and @jefflotthammer.bsky.social really did some incredible work here. I might not have had access to most of the lab GPUs for months now, but it was definitely worth it! Check out our preprint for STARLING:

020