Scientific Visualization with Grafify: Comprehensive Guide

Introduction to Grafify Scientific Plots

The grafify function in ClinicoPath provides access to the powerful grafify R package for creating beautiful, color-blind friendly scientific visualizations. This module is specifically designed for clinical research, experimental data analysis, and scientific publications.

Key Features

12+ Color-blind friendly palettes optimized for scientific visualization
Multiple plot types including scatter plots, box plots, violin plots, and specialized designs
Integrated statistical analysis with ANOVA, t-tests, correlations, and post-hoc comparisons
Experimental design support for randomized blocks, repeated measures, and factorial designs
Before-after analysis for paired data and longitudinal studies
Publication-ready styling with professional themes and customization options

Accessing Grafify in jamovi

The grafify functionality is available through the jamovi interface:

Load your data into jamovi
Navigate to Analyses → ClinicoPath → JJ Plots → Grafify
Select variables for your analysis
Choose plot type from the available options
Customize colors, themes, and statistical options
Export high-resolution plots for publication

Available Plot Types

Basic Plot Types

The grafify module supports these core visualization types:

scatterbar: Scatter plots with summary statistics and error bars
scatterbox: Box plots with individual data points overlaid
scatterviolin: Violin plots showing data distribution and points
histogram: Distribution histograms with grouping options
density: Density plots for continuous data visualization

Specialized Plot Types

befafter_box: Before-after comparisons for paired data
befafter_colors: Color-coded before-after visualizations
scatter_3d: Three-dimensional scatter plots
scatter_4d: Four-dimensional scatter plots with size/color mapping

Color Palettes

Grafify includes scientifically optimized, color-blind friendly palettes:

default: Standard palette for general use
vibrant: High-contrast colors for presentations
contrast: Maximum contrast for accessibility
bright: Bright colors for highlighting
muted: Subtle colors for professional publications
dark: Dark theme compatible colors

Example Data Setup

# Create example clinical trial data
set.seed(123)
clinical_data <- data.frame(
  patient_id = 1:120,
  treatment = rep(c("Control", "Drug A", "Drug B"), each = 40),
  age_group = rep(c("Young", "Middle", "Senior"), times = 40),
  gender = rep(c("Male", "Female"), times = 60),
  biomarker_baseline = rnorm(120, mean = 10, sd = 3),
  biomarker_followup = rnorm(120, mean = 12, sd = 4),
  response_score = rnorm(120, mean = 75, sd = 15),
  timepoint = rep(c("Baseline", "Follow-up"), each = 60)
)

# Display data structure
str(clinical_data)
head(clinical_data)

Example Visualizations

Here are examples of the types of plots you can create with grafify in jamovi:

Scatter Plot with Error Bars

# This demonstrates the type of plot grafify creates with scatterbar
summary_data <- clinical_data %>%
  group_by(treatment) %>%
  summarise(
    mean_biomarker = mean(biomarker_baseline, na.rm = TRUE),
    sd_biomarker = sd(biomarker_baseline, na.rm = TRUE),
    .groups = 'drop'
  )

ggplot(clinical_data, aes(x = treatment, y = biomarker_baseline, color = treatment)) +
  geom_point(position = position_jitter(width = 0.2), alpha = 0.7, size = 2) +
  geom_point(data = summary_data, aes(y = mean_biomarker), size = 4) +
  geom_errorbar(data = summary_data, 
                aes(y = mean_biomarker, 
                    ymin = mean_biomarker - sd_biomarker, 
                    ymax = mean_biomarker + sd_biomarker),
                width = 0.2, size = 1) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73")) +
  theme_minimal() +
  labs(title = "Biomarker Levels by Treatment Group",
       subtitle = "Points show individual values, error bars show standard deviation",
       y = "Biomarker Level (Baseline)", x = "Treatment Group") +
  theme(legend.position = "none")

Box Plot with Individual Points

ggplot(clinical_data, aes(x = treatment, y = response_score, color = gender)) +
  geom_boxplot(alpha = 0.7, outlier.shape = NA, position = position_dodge(width = 0.8)) +
  geom_point(position = position_jitterdodge(dodge.width = 0.8, jitter.width = 0.2), 
             alpha = 0.6, size = 1.5) +
  scale_color_manual(values = c("#E69F00", "#56B4E9")) +
  theme_minimal() +
  labs(title = "Response Scores by Treatment and Gender",
       subtitle = "Box plots with individual data points overlaid",
       y = "Response Score", x = "Treatment Group", color = "Gender") +
  theme(legend.position = "top")

Before-After Comparison

# Reshape data for before-after comparison
before_after_data <- data.frame(
  patient_id = rep(1:60, 2),
  treatment = rep(rep(c("Control", "Drug A", "Drug B"), each = 20), 2),
  timepoint = rep(c("Baseline", "Follow-up"), each = 60),
  biomarker = c(clinical_data$biomarker_baseline[1:60], 
                clinical_data$biomarker_followup[1:60])
)

ggplot(before_after_data, aes(x = timepoint, y = biomarker, color = treatment)) +
  geom_line(aes(group = patient_id), alpha = 0.3) +
  geom_point(alpha = 0.7, size = 2) +
  stat_summary(fun = mean, geom = "point", size = 4, shape = 15) +
  stat_summary(fun = mean, geom = "line", aes(group = treatment), size = 1.5) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73")) +
  theme_minimal() +
  labs(title = "Biomarker Changes: Baseline to Follow-up",
       subtitle = "Individual patient trajectories with group means",
       y = "Biomarker Level", x = "Time Point", color = "Treatment") +
  theme(legend.position = "top")

Statistical Analysis Integration

Grafify integrates with statistical analysis options:

Available Statistical Methods

One-way ANOVA: Comparing multiple groups
Two-way ANOVA: Factorial designs with two factors
Mixed models: Repeated measures and nested designs
T-tests: Pairwise comparisons
Correlations: Relationship analysis

Post-hoc Comparisons

Pairwise: All possible group comparisons
vs Reference: Comparisons to control group
Trends: Testing for linear/polynomial trends

Clinical Research Applications

Biomarker Analysis

Grafify is particularly useful for:

Dose-response relationships: Visualizing drug concentration effects
Treatment comparisons: Comparing multiple therapeutic interventions
Longitudinal studies: Tracking changes over time
Diagnostic performance: ROC-style visualizations

Publication-Ready Output

Features for scientific publication:

High-resolution export: 300+ DPI for journals
Customizable themes: Professional styling options
Color-blind accessibility: All palettes tested for accessibility
Statistical annotations: Automatic p-value reporting

Best Practices

Data Preparation

Clean your data before analysis
Check for outliers and missing values
Ensure proper factor levels for grouping variables
Consider transformations for skewed data

Plot Selection

Choose appropriate plot types for your data structure
Consider your audience when selecting colors
Balance detail with clarity in complex visualizations
Test accessibility with color-blind simulation tools

Statistical Considerations

Check assumptions before applying statistical tests
Consider multiple comparisons when doing post-hoc tests
Report effect sizes along with p-values
Validate findings with appropriate sample sizes

Troubleshooting

Common Issues

Missing data: Handle appropriately before analysis
Factor ordering: Ensure logical ordering of categorical variables
Color conflicts: Test palettes with your specific data
Export problems: Check resolution and file format settings

Getting Help

jamovi Community: Active user forums and support
Documentation: Comprehensive help files within jamovi
Examples: Built-in datasets and tutorials
Updates: Regular feature additions and improvements

Conclusion

The grafify module in ClinicoPath provides powerful, accessible tools for scientific visualization. Its integration with jamovi makes advanced plotting techniques available to researchers without programming experience, while maintaining the flexibility and quality needed for professional scientific publication.

By combining color-blind friendly design, statistical integration, and publication-ready output, grafify serves as a comprehensive solution for clinical and experimental data visualization needs.

Color-blind friendly plots for clinical research and experimental data

ClinicoPath

2025-07-13