← Back to Rob Beynon  |  ← Back to Dot Plot Analyzer

Protein Sequence Dot Plot Analyzer

User Manual & Reference Guide

Designed by Rob Beynon, implemented by ClaudeCode AI

1. Introduction

The Protein Sequence Dot Plot Analyzer is a powerful bioinformatics tool for comparing protein sequences and identifying conserved regions, similarities, and structural patterns. This web-based application provides both traditional dot plot visualization and modern similarity profiling with publication-quality export options.

Key Features

🧬 Multiple Scoring Methods

Choose between exact identity matching or evolutionary substitution matrices (BLOSUM62, PAM250) for biologically relevant comparisons.

🎨 Flexible Visualization

Multiple color schemes including grayscale, heat maps, and rainbow colors. Adjustable dot sizes and canvas dimensions.

📊 Similarity Profiling

Generate detailed similarity profiles showing conservation trends along sequence length with adjustable window sizes.

📁 Multi-Format Support

Load sequences from text input or FASTA files. Support for multiple sequences with all-vs-all comparison matrices.

💾 Professional Export

Export dot plots and similarity profiles as PNG, SVG, or CSV data files with intelligent naming and metadata.

⚡ Interactive Interface

Real-time parameter adjustment, sequence selection, and instant visualization updates for efficient analysis workflows.

2. Getting Started

Loading Sequences

You can input sequences in two ways:

  1. Direct Text Input: Paste sequences directly into the text areas. FASTA format is supported but not required - raw amino acid sequences work fine.
  2. File Upload: Use the file input to load FASTA files. Multiple files and multi-sequence FASTA files are supported.

💡 Pro Tip

The application comes with example sequences pre-loaded. Click "Generate Dot Plot" immediately to see how it works!

Basic Workflow

  1. Load your protein sequences (paste text or upload FASTA files)
  2. Adjust analysis parameters (window size, threshold, scoring method)
  3. Choose comparison mode (pairwise or all-vs-all)
  4. Click "Generate Dot Plot" to create visualizations
  5. Examine the dot plot and similarity profile
  6. Export results in your preferred format

3. Analysis Parameters

Parameter Range Description Recommendation
Window Size 1-20 amino acids Number of consecutive amino acids compared at each position 3-5 for detailed analysis, 10-15 for general patterns
Similarity Threshold 0-100% Minimum similarity required to display a dot 70% for BLOSUM62, 60% for PAM250, 80%+ for exact identity
Scoring Method 3 options Algorithm used to calculate similarity BLOSUM62 for most applications, PAM250 for distant relationships
Color Scheme 5 options Visual representation of similarity intensity Grayscale for publications, color schemes for presentations
Dot Size 1-10 pixels Size of similarity indicators in the plot 2-3 for detailed work, 4-6 for presentations

Scoring Methods Explained

Exact Identity

Only identical amino acids are considered similar. Fast and precise but may miss functionally similar residues like leucine/isoleucine or aspartic acid/glutamic acid that have similar biochemical properties.

Substitution Matrices: PAM vs BLOSUM

Substitution matrices are essential tools in bioinformatics that quantify the likelihood of one amino acid being substituted for another during evolution. Understanding the differences between PAM and BLOSUM matrices is crucial for choosing the right scoring method for your analysis.

PAM (Point Accepted Mutation) Matrices

📜 Historical Foundation (1978)

Development: Created by Margaret Dayhoff and colleagues, PAM matrices were the first widely-used substitution matrices. They were derived from studying closely related protein sequences (≥85% identity) to understand evolutionary relationships.

PAM250 Explained: The number "250" represents 250 PAM units of evolutionary distance. One PAM unit corresponds to 1% of amino acids being substituted. PAM250 therefore models sequences that have diverged enough for 250% of positions to have changed (with some sites changing multiple times).

Construction Method: Built from phylogenetic trees of closely related sequences, then extrapolated mathematically to model more distant relationships. This extrapolation assumes evolutionary rates remain constant over time.

BLOSUM (BLOcks SUbstitution Matrix) Matrices

🧬 Modern Approach (1992)

Development: Created by Steven and Jorja Henikoff, BLOSUM matrices addressed limitations of PAM by using a different approach. Instead of extrapolation, they directly analyzed conserved regions (blocks) from distantly related proteins.

BLOSUM62 Explained: The number "62" represents 62% sequence identity clustering threshold. Sequences more than 62% identical were clustered together to remove bias, ensuring the matrix reflects substitutions across a broader evolutionary range.

Construction Method: Based on direct observation of substitutions in conserved protein blocks from the BLOCKS database. No mathematical extrapolation required, making it more empirically grounded.

Key Differences Summary

Aspect PAM250 BLOSUM62
Data Source Closely related sequences (≥85% identity) Distantly related sequences (≤62% identity)
Construction Mathematical extrapolation from close relationships Direct observation of distant relationships
Best For Distant evolutionary relationships Moderate evolutionary distances
Conservative Substitutions More generous scoring for similar amino acids More stringent, requires stronger similarity
Sensitivity Higher sensitivity, may detect more distant homologs Lower false positives, more specific matches
Gap Penalties Typically uses higher gap penalties Typically uses lower gap penalties

Practical Usage Guidelines

🔍 Use BLOSUM62 When:

  • Comparing sequences with moderate similarity (30-80% identity)
  • Looking for functional domains or motifs
  • General protein similarity searches
  • You want fewer false positives
  • Working with well-characterized protein families

🧭 Use PAM250 When:

  • Searching for very distant evolutionary relationships
  • Comparing highly divergent sequences (<30% identity)
  • Phylogenetic analysis of ancient relationships
  • You want maximum sensitivity to detect remote homologs
  • Studying protein superfamilies or ancient domains

💡 Matrix Selection Strategy

Start with BLOSUM62 for most analyses - it's the current standard and works well for the majority of protein comparisons. Switch to PAM250 only when you specifically need to detect very distant relationships or when BLOSUM62 produces too few matches.

⚠️ Important Considerations

Threshold Adjustment: When switching between matrices, adjust your similarity threshold accordingly. PAM250 typically requires lower thresholds (50-70%) while BLOSUM62 works well with higher thresholds (70-85%) due to their different scoring philosophies.

Color Schemes

Black
Grayscale
Red Heat
Blue Scale
Rainbow

4. Comparison Modes

Pairwise Comparison

Compare two specific sequences with detailed visualization including:

  • Full-resolution dot plot with axis labeling
  • Similarity profile graph showing trends along sequence length
  • Detailed statistics and analysis parameters
  • Independent export options for both visualizations

💡 Reading Dot Plots

Diagonal lines indicate conserved regions. The main diagonal (bottom-left to top-right) shows perfect alignment. Parallel diagonals suggest repeats or similar domains.

All vs All Matrix

Compare multiple sequences simultaneously with a comparison matrix showing:

  • Mini dot plots for each sequence pair
  • Diagonal cells showing sequence names
  • Symmetric matrix layout for easy pattern recognition
  • Consistent scaling across all comparisons

⚠️ Performance Note

All-vs-all comparisons can be computationally intensive. For large numbers of sequences (>10), consider using smaller window sizes or higher thresholds.

5. Similarity Profile

The similarity profile provides a complementary view to dot plots by showing similarity trends along the sequence length as a line graph.

Profile Features

  • Adjustable Window: Choose from 1, 5, 10, or 20 amino acid windows
  • Statistical Summary: Average, maximum, minimum similarity, and high-similarity regions
  • Visual Trends: Peaks indicate conserved regions, valleys show variable areas
  • Independent Export: Save profiles separately from dot plots

Interpreting Profiles

High Peaks (>80%)

Highly conserved regions, likely functionally important domains or structural elements.

Sustained Plateaus

Moderately conserved regions, possibly indicating structural domains or functional motifs.

Low Valleys (<40%)

Variable regions, often corresponding to loops, linkers, or species-specific adaptations.

6. Export Options

The application provides comprehensive export capabilities for both visualizations and underlying data:

Image Formats

Format Best For Features
PNG Presentations, web display High quality, universal compatibility, small file size
SVG Publications, print media Vector format, infinite scaling, professional quality

Data Export

CSV export includes:

  • Complete analysis metadata (parameters, sequences, date)
  • Position coordinates for all similarity matches
  • Similarity scores and sequence windows
  • Ready for statistical analysis or further processing
# Example CSV output format: # Generated: 2025-01-08T10:30:00.000Z # X-axis sequence: Protein_A # Y-axis sequence: Protein_B # Scoring method: blosum62 X_Position,Y_Position,Similarity_Score,X_Sequence_Window,Y_Sequence_Window 0,0,85.50,MKTV,MKTV 1,1,92.33,KTVR,KTVR ...

7. Tips & Best Practices

Optimization Strategies

Window Size Selection

Small windows (1-3) for finding short motifs, medium windows (5-10) for domains, large windows (15-20) for overall similarity trends.

Threshold Tuning

Start with default settings, then adjust threshold to balance sensitivity (lower values) vs. specificity (higher values).

Scoring Method Choice

Use BLOSUM62 for general comparisons, PAM250 for distant evolutionary relationships, exact identity for finding perfect matches.

Visualization Tips

Use grayscale for publications, color schemes for presentations. Adjust dot size based on image resolution and intended use.

💡 Performance Tip

For very long sequences (>1000 amino acids), consider using larger window sizes (10-20) and higher thresholds (75-85%) to reduce computation time while maintaining biological relevance.

Common Use Cases

  • Domain Analysis: Use medium windows (5-10) with BLOSUM62 to identify conserved protein domains
  • Motif Discovery: Small windows (1-3) with exact identity to find short conserved sequences
  • Evolutionary Studies: PAM250 scoring with moderate thresholds to trace distant relationships
  • Quality Control: High thresholds (90%+) to verify sequence identity or find sequencing errors

8. Technical Details

Algorithm Overview

The application uses a sliding window approach where each position in the sequence comparison generates a similarity score based on the selected scoring method. Dots are placed where similarity exceeds the specified threshold.

Coordinate System

  • Origin (0,0) is positioned at the bottom-left corner
  • X-axis represents the first sequence (horizontal)
  • Y-axis represents the second sequence (vertical)
  • Position numbers indicate amino acid indices starting from 0

Browser Compatibility

The application works in all modern web browsers with HTML5 Canvas support. No plugins or additional software required.

⚠️ Data Privacy

All sequence analysis is performed locally in your browser. No data is transmitted to external servers, ensuring complete privacy of your sequences.

File Format Support

  • FASTA format (.fasta, .fas, .fa)
  • Plain text files (.txt)
  • Multiple sequences per file
  • Standard amino acid codes (20 standard + B, Z, X)

For support or questions about this application, please refer to the original developer.

Manual version 1.0 - January 2025