NC Bioinformatics Figure Skills

NC-style bioinformatics figure training, code mining, and R plotting toolkit collection for Codex.

What this project is for

This repository collects a Codex skill, reference notes, and small R plotting packages for learning and building Nature Communications-style bioinformatics figures.

It is useful when you want to:

• decide what belongs in a single-cell, spatial, trajectory, communication, benchmark, or multi-omics main figure;
• inspect GitHub-backed paper code and turn figure patterns into reusable plotting recipes;
• route a project to the right local plotting package;
• build source-data-first R figure panels that can be audited and reused.

Contents

skills/nc-bioinformatics-figure-skills/
packages/ncfigR/
packages/scfigR/
packages/spfigR/
packages/commfigR/
packages/trajfigR/
packages/benchfigR/
packages/multiomfigR/

R package map

Package	Scope
ncfigR	Shared base toolkit for embedding, composition, heatmap, network, trajectory, benchmark, and export panels
scfigR	Single-cell atlas overview figures
spfigR	Spatial tissue niche and spatial-feature figures
commfigR	Cell-cell communication, ligand-receptor, and differential communication figures
trajfigR	Pseudotime, branch, gene-trend, velocity, and state-transition figures
benchfigR	Benchmark plus biological case-study figures
multiomfigR	Multi-omics integration, modality metric, feature-link, and pathway-program figures

Codex skill

The skill entry point is:

skills/nc-bioinformatics-figure-skills/SKILL.md

The reference folder includes figure decision rules, visual QA scorecards, code-pattern inventories, package API notes, project scaffold templates, and reusable plotting recipe cards.

Install package locally

Example:

install.packages("remotes")
remotes::install_local("packages/ncfigR")
remotes::install_local("packages/scfigR")

Install ncfigR first because the other packages reuse its theme, palette, export, and base-panel helpers.

License

MIT License.

Paper titles, article links, and GitHub/code links are used as learning indexes. Follow the license of each original paper, dataset, and code repository when reproducing specific figures.