Discovery only matters once it's communicated. The last mile of research (turning a result into a figure a reviewer trusts and a reader understands) is where a surprising amount of a scientist's time goes, and where a surprising number of tools fall short. We've been investing in that last mile.
Two of the workhorses of differential-expression reporting, the volcano plot and the MA plot, are now fully interactive to edit inside Rosalind. Adjust significance cutoffs and thresholds, add and reposition the labels that matter, and tune colors until the figure says exactly what your data says. When it's right, export it publication-ready. No exporting a table, no rebuilding the plot in another tool, no losing the thread between the analysis and the image.
A good figure isn't only beautiful. It's defensible. That's why the plot work sits alongside our Publication module, which tracks figure exports and analysis provenance as you assemble a manuscript. The result is a tighter loop between the analysis you ran and the figure you publish, with a record of how you got there.
This matters beyond convenience. Across the industry, funders, journals, and partners are asking harder questions about reproducibility and how a result was produced. Provenance is quietly becoming table stakes. Our bet is that the platforms scientists trust will be the ones where rigor is built in, not bolted on, where the path from raw data to published figure is transparent by default.
Making that path shorter and clearer is one of the most direct ways we can give researchers their time back. It's also, we think, simply how good science tools should work. And it's foundational in a deeper sense: a platform that faithfully records how every result was made is a platform you can eventually trust to help produce and interpret results: the bedrock any credible, AI-assisted science will have to stand on. There's more coming here. This is a foundation we intend to keep building on, deliberately, for years.