Protein Family Phylogenetic Analysis

Complete workflow for protein family evolutionary analysis: quality control → conservation → coevolution → phylogeny → publication report.

Quick Start

Input: FASTA file with protein sequences (any family, any size)
Output: Publication-ready report with phylogenetic tree, conservation analysis, coevolution networks, and high-quality figures

Typical workflow:

# 1. Quality control (removes low-quality sequences)
bash scripts/01_quality_control.sh input.fasta output_dir/

# 2. Conservation analysis
bash scripts/02_conservation.sh output_dir/qc/final.fasta output_dir/

# 3. Coevolution analysis
bash scripts/03_coevolution.sh output_dir/qc/final.fasta output_dir/

# 4. Phylogenetic tree
bash scripts/04_phylogeny.sh output_dir/qc/final.fasta output_dir/

# 5. Generate figures
bash scripts/05_visualize.sh output_dir/

# 6. Create report
bash scripts/06_report.sh output_dir/ "Family Name"

Workflow Overview

Stage 1: Quality Control (references/01-quality-control.md)

Purpose: Filter raw sequences to high-quality, non-redundant dataset

Steps:

1. Literature validation (remove predicted sequences)
2. Length filtering (remove fragments/fusions)
3. CD-HIT redundancy removal (90% identity)
4. Complexity check (remove low-complexity regions)
5. Motif validation (confirm family membership)
6. MAFFT alignment (high accuracy mode)
7. trimAl trimming (automatic strategy)
8. Final validation (gap ratio, coverage)

Key parameters:

• CD-HIT threshold: 90% (adjustable 70-95%)
• Length range: mean ± 2 SD
• Gap threshold: < 30% per position
• Motif coverage: > 50%

Output: qc/final.fasta (high-quality aligned sequences)

Stage 2: Conservation Analysis (references/02-conservation.md)

Purpose: Identify functionally important conserved residues

Method: Shannon entropy

• H_norm < 0.3: Highly conserved
• H_norm 0.3-0.6: Moderately conserved
• H_norm > 0.6: Variable

Output:

• Conserved positions list
• Conservation landscape plot
• Gap vs conservation scatter plot

Stage 3: Coevolution Analysis (references/03-coevolution.md)

Purpose: Identify residue pairs that evolve together

Method: Normalized Mutual Information (NMI)

• Corrects for phylogenetic bias
• Identifies structural/functional coupling
• Builds coevolution network

Output:

• Coevolved position pairs (MI scores)
• Network graph (hub identification)
• Hub residue heatmap

Stage 4: Phylogenetic Analysis (references/04-phylogeny.md)

Purpose: Reconstruct evolutionary relationships

Method: IQ-TREE maximum likelihood

• Automatic model selection (ModelFinder)
• UFBoot2 ultrafast bootstrap (1000 replicates)
• Convergence check (> 0.99 required)

Output:

• Phylogenetic tree (.treefile)
• Bootstrap consensus tree (.contree)
• Model parameters (.iqtree)

Stage 5: Visualization (references/05-visualization.md)

Purpose: Generate publication-quality figures (300 DPI)