Quarto & Presentation Automation Prompts

This guide gathers prompts and workflows for automatically generating presentations from analysis pipelines and documentation, covering both web-based Reveal.js slide decks and Microsoft PowerPoint (.pptx) presentations.

1. Web-Based Reveal.js Presentations with Quarto

This workflow transforms existing documentation, PDF publications, or markdown notes into a styled, interactive Reveal.js presentation with a modern aesthetic, calibrated typography, and self-contained asset embedding.

Prompts and Refinements

Initial Generation Prompt:

“Create a quarto slide deck from [slidedeck].md (or article PDF in data/) in light mode and place it in folder quarto/ as [slidedeck].qmd. Start by developing a staged outline based on the main sections. Adjust slide styling to use a compact layout so that content fits neatly on each slide. Do not use chalkboard features; instead embed all resources (embed-resources: true). Calibrate inline backtick code elements so their font size matches the visual height of surrounding text. Update .gitignore for quarto documents as needed.”

Production-Grade Refinement Directives:

  • Typography & Back-Quoted Strings: Calibrate inline monospace elements (code:not(pre code)) to 0.88em with baseline alignment and inherited line height across paragraphs, list items, tables, and spans to match adjacent sans-serif font size without distorting vertical rhythm.
  • Figure Asset Sourcing: Link directly to official CDN or publication URLs for high-resolution figures when available; Quarto dynamically base64-encodes them into the self-contained deck, avoiding large local binary files in version control.
  • Layout Architecture & Collision Prevention: For complex conceptual overviews or multi-part hypotheses, use responsive 2×2 callout grids (::: {.columns}) or statistical cards (.stat-card) rather than tall SVG/Mermaid flowcharts to prevent browser clipping and hidden slide rendering collisions.
  • Slide Navigation & Overview: Configure 16:9 widescreen slides with progress bars, slide numbers, and matrix overview mode (O key).

Key Configurations & Structure

  1. Quarto Slide Deck (quarto/[slidedeck].qmd):
    • Web-based Reveal.js format with self-contained bundling: 
      format:
  revealjs:
    theme: [default, custom.scss]
    slide-number: true
    embed-resources: true
    width: 1600
    height: 900
    transition: fade
    • Structured into logical sections using background-colored slide breaks (# Section Title {background-color="..."}).
    • Utilizes advanced Quarto features such as column layouts (::: {.columns}), callout boxes, and incremental lists (::: {.incremental}).
    • Uses responsive 2×2 card or callout grids for multi-part hypotheses or methods to prevent SVG rendering collisions.
    • Adjusts all relative file links to match the subdirectory structure.
  2. Custom Presentation Styling (quarto/custom.scss):
    • Light-Mode Scientific Aesthetic: Clean Slate-50 background (#f8fafc), Slate-900 headings (#0f172a), Sky-600 links/accents (#0284c7), and subtle slate borders (#e2e8f0).
    • Modern Font Pairing: Google Fonts pairing with Outfit for presentation and slide headings, and Inter for body copy, tables, and statistics.
    • Back-Quoted String Calibration (Inline Code): Monospace fonts inherently have taller x-heights and cap-heights than proportional sans-serif fonts. Calibrate inline backtick code elements to match surrounding text: 
      // Inline code styling (backticks) - calibrated to match adjacent text size
code:not(pre code) {
  font-family: var(--r-code-font, ui-monospace, SFMono-Regular, Menlo, Monaco, Consolas, "Liberation Mono", monospace);
  font-size: 0.88em !important;
  background-color: #f1f5f9;
  padding: 0.06em 0.28em;
  border-radius: 4px;
  color: #0369a1;
  border: 1px solid #e2e8f0;
  vertical-align: baseline;
  line-height: inherit;
  font-weight: 500;
}

li code:not(pre code),
p code:not(pre code),
td code:not(pre code),
th code:not(pre code),
span code:not(pre code) {
  font-size: 0.88em !important;
}
    • Card & Compact Slide Utilities: Includes .stat-card for KPI/metric highlights and .compact-slide class for high-density slides.
  3. Rendered Output, Verification & PDF Export:
    • Compile HTML: Compile using quarto render quarto/[slidedeck].qmd (verifying exit code 0 and clean asset embedding).
    • Presentation Navigation:
      • F: Toggle full screen
      • O: Toggle slide overview matrix
      • Space / Arrow Keys: Advance slides
    • Export to PDF (Headless Chrome — 11” × 8.5” Landscape): 
      "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" \
  --headless \
  --run-all-compositor-stages-before-draw \
  --no-pdf-header-footer \
  --print-to-pdf="quarto/[slidedeck].pdf" \
  "file://$(pwd)/quarto/[slidedeck].html?print-pdf"
    • Export to PDF (Browser Print Dialog): Open quarto/[slidedeck].html?print-pdf in Chrome/Safari/Edge, press Cmd + P, select Save as PDF, Layout: Landscape, Margins: None, and check Background graphics.

2. Automated PowerPoint Deck Generation from R Pipelines

This workflow automates the assembly of a Microsoft PowerPoint presentation (.pptx) directly from an R data analysis pipeline, embedding generated PNG figures, titles, tables of contents, and biological narrative slides.

Case Study: Glucose Tolerance Test (GTT) Analysis Pipeline

Date: 2026-03-19 (Updated: 2026-03-25)

Execution Prompt Used:

“Run the prompt and do all the steps listed in promptsAlan.md. Make sure all software needed is installed.”

Consolidated Pipeline Instructions

  1. Data Initialization:

    Analyze Glucose Tolerance Test (GTT) data from the master spreadsheet 'NDNB-25 B6 and NZO Masterdate sheet v2 R.xlsx'.

  2. Time-Course Line Graphs (Glucose, Insulin, & Log Ratio):

    Generate time-course line graphs for GTT data at 0, 15, 30, and 120 minutes vs. time. Create separate plots for:

    • Glucose levels
    • Insulin levels
    • log(Insulin / Glucose) ratio

    Output individual figures for both the NZO and B6 strains across each of the three metrics (6 plots total). Map Vehicle-treated to black dots and lines, and Compound-treated to blue dots and lines. Include standard error bars.

  3. Cross-Sectional Scatter Plot:

    Create a cross-sectional scatter plot mapping change in fasting glucose (Y-axis) against change in body weight (X-axis) across all mice.

    • Color NZO strain blue and B6 strain black.
    • Drug-treated mice as circles, vehicle-treated as squares.
    • Add linear regression lines for each of the 4 subgroups.
    • Calculate and display Pearson correlation coefficients (r) and p-values on the plot.
  4. File Management:

    Format spaces to underscores (Hsp90_GTT.R), and explicitly add the .R extension.

  5. Presentation Generation via Quarto:

    Use Quarto to automatically generate a PowerPoint deck (.pptx) embedding all of the outputted PNG figures. Ensure the presentation includes:

    • A two-column Table of Contents slide with functional intra-document hyperlinks.
    • A background section on Gluconeogenesis, split across multiple clean slides with academic references.
  6. Insulin Baseline Normalization:

    Remove time 0 insulin baseline measurements and calculate two new subgroup metrics: the Insulin / Insulin_0 ratio and log-regression residuals (log(Insulin) ~ log(Insulin_0)). Plot both over time, splitting NZO and B6 strains side-by-side using facet_wrap. Add these plots to the PowerPoint deck.

This site uses Just the Docs, a documentation theme for Jekyll.