BIOSZEN

Biological data analysis and visualization in R/Shiny
From data import to reproducible plots, statistics, metadata, and bundle exports.

IMPORTANT: For the most complete workflow, use Platemap + Curves input mode (Datos + PlotSettings + well-based curves).

TIP: If your dataset is large, start with .csv in Load Data and keep metadata as .xlsx.

• Why BIOSZEN
• Core Capabilities
• Quick Start
• Choose the Right Input Mode
• Recommended Workflow
• Visual Gallery
• Common User Scenarios
• Documentation
• AI Skill for Input Preparation
• Troubleshooting
• Citation
• License

BIOSZEN is an R package with a modular Shiny app focused on biological experiment analysis. It helps you go from raw data to publication-ready visualizations while preserving reproducibility through metadata and bundle exports.

• Plot families: Boxplot, Barplot, Violin, Curves, Stacked, Correlation, Heatmap, Correlation Matrix.
• Statistical workflows: normality checks, significance testing, post hoc comparisons, p-value correction.
• Control-based normalization with replicate-aware behavior.
• Biological and technical replicate QC (manual + automatic strategies).
• Composition panel for multi-plot layouts and export (PNG, PDF, PPTX).
• Reproducibility features: metadata export/import and ZIP bundles.
• Growth module with extracted metrics: uMax, doub_time, lag_time, AUC, ODmax, and related fields.

1. Install R (>= 4.1).
2. Open this project folder.
3. Run one of the following:

Rscript app.R

Or inside R:

BIOSZEN::run_app()

4. Open the local URL shown in the console.

NOTE: On first launch, dependencies may be installed into a local R_libs directory. Keep this folder to avoid reinstalling packages.

• Platemap + Curves (recommended)
  Use it when: You need full plotting + stats + replicate/QC workflows.
  Required files: Datos + PlotSettings workbook, plus curves file (Time + wells).

• Grouped parameters
  Use it when: You only need parameter plots/statistics.
  Required files: Grouped parameter workbook in Load Data.

• Summary (Mean/SD/N)
  Use it when: Raw replicate rows are not available.
  Required files: Summary workbook with dedicated summary sheets.

• CSV mode
  Use it when: Dataset is large and you want lighter IO.
  Required files: .csv for data and/or curves; metadata still .xlsx.

1. Load primary file in Load Data.
2. Optionally load curves in Load Curves.
3. Apply filters (group/media/replicate).
4. Choose plot type and visual settings.
5. Optionally normalize by control.
6. Run statistics and significance annotation.
7. Export plot/data/statistics/metadata/bundle.

CAUTION: If curves do not align with your groups, verify the Well mapping in platemap and the curves column names (A1, A2, etc.).

• Use Summary (Mean/SD/N) mode.
• Focus on parameter visualization and supported statistical paths.
• Expect some normality/non-parametric options to be limited if raw observations are unavailable.

• Use BiologicalReplicate and TechnicalReplicate columns.
• Start with manual include/exclude.
• Then apply IQR filtering and Keep-N reproducibility filters.

• Export metadata after configuring your analysis.
• Export bundle ZIP at the end.
• Re-import metadata in a new session to restore state.

• Main English manual: inst/app/www/MANUAL_EN.md
• Main Spanish manual: inst/app/www/MANUAL_ES.md
• PDF manuals:
  • inst/app/www/MANUAL_EN.pdf
  • inst/app/www/MANUAL_ES.pdf
• Input templates: inst/app/www/reference_files/

Reference templates include:

• Ejemplo_platemap_parametros.xlsx
• Ejemplo_curvas.xlsx
• Ejemplo_parametros_agrupados.xlsx
• Ejemplo_input_summary_mean_sd.xlsx

This repository includes an optional reusable AI-agent skill in the skills/bioszen-platemap-curves/ GitHub folder. To use it, give that GitHub folder URL to the corresponding AI tool or agent so it can read or acquire the skill. If the agent needs local files, download the repository ZIP from https://github.com/bioszen/BIOSZEN/archive/refs/heads/main.zip and copy the skills/bioszen-platemap-curves/ folder into your agent's skill system.

The skill can be used from Codex, Claude, Antigravity, or similar agentic coding tools to create, repair, and validate BIOSZEN-compatible platemap and curves workbooks before upload. It is intended for arbitrary numeric plotting datasets: it does not assume fixed parameter names, fixed experiment labels, or a specific measurement type.

Use it when you need help generating a Datos + PlotSettings platemap from any readable source file with data, creating or correcting a separate curves workbook, repairing stale PlotSettings, fixing parameter-name typing mistakes that make BIOSZEN reject or misread a platemap, or checking that Datos$Well matches curve columns exactly. The skill is documentation/tooling only and does not change the BIOSZEN Shiny application.

• Rscript not found
  Quick check: Install R and ensure Rscript is on PATH (or run with full executable path).

• Upload rejected
  Quick check: Verify required sheet names and column names.

• No plot shown
  Quick check: Confirm selected parameter/group still exists after filters.

• Stats unavailable
  Quick check: Check whether your input mode supports the selected statistical route.

• Slow performance
  Quick check: Reduce selected parameters and disable heavy overlays.

Szenfeld, B. (2026). BIOSZEN (2.0.2). Zenodo. https://doi.org/10.5281/zenodo.18217522

GPL-3.0