Understanding Phylogenetic Analysis

From DNA Sequences to Evolutionary Trees

ENTM201L - Lab Theory

Why Build Phylogenetic Trees?

After Sanger sequencing reveals your mosquito COI sequences in, you possess genetic barcodes that identify species. But DNA sequences contain far more information than just species names. They record evolutionary history - the branching pattern of descent from common ancestors.

Phylogenetic trees (also called phylogenies or evolutionary trees) visualize these relationships, allowing you to:

Trace invasion routes of Aedes albopictus across continents
Identify sister species (closest evolutionary relatives)
Estimate divergence times (when populations separated)
Detect cryptic species (hidden genetic diversity)
Understand pathogen transmission (which mosquito lineages spread diseases)
Inform conservation (identify evolutionary distinct populations for protection)

In ENTM201L, you will build a phylogenetic tree from your class COI sequences plus reference sequences from GenBank, using industry-standard software (MAFFT for alignment, IQ-TREE for tree building). This connects your lab work to real-world applications in vector surveillance, invasion biology, and evolutionary research.

What is a Phylogenetic Tree?

Tree Anatomy and Terminology

A phylogenetic tree is a branching diagram showing evolutionary relationships among organisms or genes.

Basic components:

 ┌─── Aedes albopictus (USA)
 ┌─────┤
 │ └─── Aedes albopictus (Japan)
 ┌─────┤
 │ │ ┌─── Aedes aegypti (Brazil)
 │ └─────┤
 ────┤ └─── Aedes aegypti (Florida)
 │
 │ ┌───────── Culex pipiens
 └─────┤
 └───────── Culex tarsalis

Key terms:

Term	Definition	Example
Tips (leaves)	Terminal nodes representing species/samples	Aedes albopictus (USA)
Branches	Lines connecting nodes, represent evolutionary change	Horizontal/diagonal lines
Nodes (internal)	Branching points representing common ancestors	Where Ae. albopictus USA & Japan connect
Root	Base of tree, most recent common ancestor of all tips	Leftmost point where all branches connect
Clade	Group of organisms sharing common ancestor	All Aedes species = one clade
Sister taxa	Two closest relatives sharing immediate common ancestor	Ae. albopictus USA & Japan
Outgroup	Distantly related species used to root tree	Culex (if analyzing Aedes)

Rooted vs. Unrooted Trees

Rooted tree:

Has a root showing direction of evolution (time flows from root to tips)
Can infer ancestral states
Shows which lineages are older vs. newer

Unrooted tree:

Shows relationships without direction
Cannot tell which species is ancestral
Requires outgroup to convert to rooted tree

For ENTM201L: We will build rooted trees using Culex or Anopheles as outgroup to root Aedes trees.

Branch Lengths: Evolutionary Distance

Branch length represents amount of evolutionary change (genetic divergence). Measured as:

Substitutions per site: Average number of nucleotide changes per position

- 0.01 = 1% divergence (1 substitution per 100 bp on average)

- 0.10 = 10% divergence

In COI phylogenies:

- Within species: 0.001-0.02 (0.1-2% divergence)

- Between species (same genus): 0.05-0.15 (5-15%)

- Between genera: 0.20-0.30 (20-30%)

Example interpretation:

Ae. albopictus USA and Japan differ by 0.5% → diverged recently (within last 50,000-100,000 years, roughly)
Ae. albopictus and Ae. aegypti differ by 12% → diverged ~50 million years ago

Multiple Sequence Alignment: The Foundation

Why Alignment is Essential

Problem: You have 20 COI sequences of varying lengths (650-700 bp each). To compare them, you need to know which positions are homologous (descended from same ancestral base). Without alignment:

Sequence 1: ATCGATCGAA...
Sequence 2: ATCGTCGAA... (deleted one A)
Sequence 3: ATCGAATCGAA... (inserted one A)

You cannot tell position 5 in Seq 1 corresponds to position 4 in Seq 2 or position 6 in Seq 3.

With alignment (inserts gaps to maintain positional homology):

Sequence 1: ATCGATCGAA
Sequence 2: ATCG-TCGAA (gap shows deletion)
Sequence 3: ATCGAATCGAA (insertion)
 ||||*||||||
 Positions now comparable

Alignment principle: Insert gaps (indels) to maximize similarity across sequences, ensuring each column represents one ancestral site.

How MAFFT Works

MAFFT (Multiple Alignment using Fast Fourier Transform) is the most widely used alignment software for nucleotide and protein sequences. Algorithm overview:

1. Pairwise distance calculation: Compare all sequences to each other

2. Guide tree construction: Build rough tree showing approximate relationships

3. Progressive alignment: Align most similar sequences first, gradually add more distant sequences

4. Refinement: Iteratively improve alignment by realigning subgroups

Why MAFFT for COI:

Fast: Aligns 100 sequences in seconds
Accurate: Handles both conserved and variable regions
Codon-aware: Can align protein-coding genes respecting reading frame
Free and open-source: Available for Mac, Windows, Linux

Command-line usage (if using terminal):

mafft --auto input.fasta > aligned.fasta

Alignment strategies:

--auto: MAFFT selects best algorithm based on dataset size (recommended)
--localpair: More accurate for divergent sequences (slower)
--retree 2: Fast, good for large datasets

Alignment Quality: What to Check

High-quality alignment:

Conserved regions: Blocks of identical bases across all sequences
Few gaps: Gaps only where indels are biologically plausible
Reading frame preserved: For protein-coding genes like COI, no frameshift mutations (which would create stop codons)

Poor-quality alignment:

Excessive gaps: Many columns with >50% gaps (over-alignment)
Shifted regions: Blocks of sequence clearly misaligned
Frameshifts: Reading frame disrupted (suggests NUMTs or sequencing errors)

Inspecting alignment (in Geneious or MEGA):

1. Translate to protein: COI codes for Cytochrome Oxidase protein

- Should have no stop codons (except terminal TAA)

- Conserved amino acids (e.g., active site residues) should align

2. Check for NUMTs: Nuclear mitochondrial pseudogenes often have stop codons or frameshifts

- If present, exclude from phylogeny

3. Trim poorly aligned regions: Remove first/last 10-20 bp if ambiguous

Phylogenetic Inference Methods

Distance-Based Methods (UPGMA, Neighbor-Joining)

Principle: Calculate genetic distance between all pairs of sequences, cluster most similar pairs iteratively. UPGMA (Unweighted Pair Group Method with Arithmetic Mean):

Assumes constant evolutionary rate (molecular clock)
Builds ultrametric tree (all tips equidistant from root)
Fast but often inaccurate (clock assumption violated in real data)

Neighbor-Joining:

Does not assume molecular clock
Minimizes total branch length
Faster than likelihood methods
Good for: Quick exploratory trees, large datasets (>1000 sequences)

Limitations:

Ignores alignment details (only uses pairwise distances)
Cannot incorporate complex substitution models
Less accurate than likelihood methods

Maximum Likelihood (ML): The Gold Standard

Principle: Find the tree (topology + branch lengths) that maximizes the probability of observing your sequence data, given a substitution model. How it works:

1. Propose a tree topology

2. Calculate likelihood: P(data | tree, model)

3. Optimize branch lengths to maximize likelihood

4. Try different topologies, keep best one

5. Repeat until convergence

Why ML is superior:

Model-based: Uses realistic substitution models (see below)
Statistically rigorous: Can calculate confidence (bootstrap support)
Handles rate variation: Accounts for different substitution rates across sites
Accurate: Outperforms distance and parsimony methods in simulations

Software:

IQ-TREE (ENTM201L uses this): Fastest ML software, automatic model selection
RAxML: Popular alternative, good for very large datasets
PhyML: Older but still widely used

Bayesian Inference: Posterior Probabilities

Principle: Calculate posterior probability of each tree given data, using Bayesian statistics. Advantages over ML:

Provides posterior probabilities for nodes (more intuitive than bootstrap)
Can incorporate prior information (e.g., fossil calibrations for divergence dating)
Samples many high-probability trees (accounts for uncertainty)

Disadvantages:

Much slower than ML (hours to days for 100 sequences)
Requires careful interpretation of convergence diagnostics
More complex to set up

Software: MrBayes, BEAST (for divergence dating) For ENTM201L: We use ML (IQ-TREE) because it is fast, accurate, and easier to run than Bayesian methods.

Substitution Models: How DNA Evolves

Why Models Matter

Not all nucleotide changes are equally likely. Transitions (A↔G, C↔T) occur more frequently than transversions (A↔C, A↔T, G↔C, G↔T) due to:

Chemical similarity (purines A/G vs. pyrimidines C/T)
DNA repair mechanisms (transition errors less often corrected)

Substitution models describe how nucleotides change over time, allowing accurate estimation of evolutionary distances and tree topologies.

Common Models

Jukes-Cantor (JC69) - Simplest model:

All substitutions equally likely
Equal base frequencies (A=T=G=C=25%)
Rarely accurate for real data

Kimura 2-parameter (K2P) - Transition/transversion distinction:

Transitions (A↔G, C↔T) at rate κ
Transversions (all others) at rate 1
Still assumes equal base frequencies
Common for COI barcoding (simple, reasonably accurate)

General Time-Reversible (GTR) - Most complex:

Six different substitution rates (A↔C, A↔G, A↔T, C↔G, C↔T, G↔T)
Unequal base frequencies
Most flexible, best fit for most datasets

Rate heterogeneity (+I+G):

+I: Some sites invariant (never change, e.g., functionally critical positions)
+G: Gamma distribution of rates across sites (some fast, some slow)
GTR+I+G: Most comprehensive model, often best for protein-coding genes

Model Selection

IQ-TREE automatic model selection:

Tests all models (88 common models)
Calculates fit using BIC (Bayesian Information Criterion) or AIC (Akaike Information Criterion)
Selects best-fit model
Runs this automatically when you use -m MFP flag

Typical result for COI:

Best-fit model: TN+F+I+G4
 TN = Tamura-Nei (different transition rates for A↔G vs. C↔T)
 +F = Unequal base frequencies
 +I = Invariant sites
 +G4 = Gamma distribution of rates (4 categories)

Base frequencies:
 A=29.8%, C=15.4%, G=16.1%, T=38.7% (AT-rich, typical for insect mtDNA)

Bootstrap Values: Measuring Confidence

What Bootstrap Measures

Problem: How confident are we in each node (clade) on the tree? Bootstrap resampling (introduced by Felsenstein, 1985) tests tree robustness:

1. Take your alignment (e.g., 650 bp)

2. Resample with replacement: Randomly select 650 positions (some sampled multiple times, some not at all)

3. Build tree from resampled alignment

4. Repeat 1,000 times

5. Count how often each clade appears in the 1,000 bootstrap trees

6. Bootstrap value = % of trees containing that clade

Interpretation:

≥95%: Strong support, clade very reliable
80-95%: Moderate support, likely correct
70-80%: Weak support, uncertain
<70%: Poor support, clade unreliable (may be artifact)

Example:

 ┌─── Ae. albopictus (USA)
 ┌─────┤ 99%
 │ └─── Ae. albopictus (Japan)
 ┌─────┤
 │ │ ┌─── Ae. aegypti (Brazil)
 ────┤ └─────┤ 95%
 │ 85% └─── Ae. aegypti (Florida)
 │
 └───────────── Culex pipiens (outgroup)

Interpretation:

99% support: Ae. albopictus USA and Japan are definitely sister lineages
95% support: Ae. aegypti Brazil and Florida cluster together (high confidence)
85% support: Aedes monophyly (all Aedes share common ancestor excluding Culex) is moderately supported

UFBoot: Ultrafast Bootstrap

Standard bootstrap: Slow (rebuilds tree 1,000 times)

For 100 sequences, 30 minutes per tree × 1,000 = 500 hours

Ultrafast bootstrap (UFBoot) in IQ-TREE:

Approximates standard bootstrap with clever resampling
10-100× faster
Equally accurate (tested extensively)
Default in IQ-TREE when you use -bb 1000

Result: 100-sequence tree with 1,000 bootstrap replicates in 10-20 minutes instead of days.

Reading and Interpreting Trees

Topology: Who is Related to Whom?

Monophyletic group (clade): All descendants of a common ancestor

Example: All Aedes species (if they cluster together)

Paraphyletic group: Some but not all descendants of common ancestor

Example: "Aedes" if Ochlerotatus is included (because Ochlerotatus is nested within Aedes)

Polyphyletic group: Members do not share recent common ancestor

Example: "Tree-hole breeding mosquitoes" (many unrelated lineages independently evolved this habit)

Tree reading principle: Only monophyletic groups are evolutionarily meaningful. Paraphyletic and polyphyletic groups are artificial.

Rotating Branches: Trees Have No Intrinsic Order

Critical concept: You can rotate branches around any node without changing relationships.

These three trees are identical:

Tree 1: Tree 2: Tree 3:
 ┌─── A ┌─── B ┌─── C
────┤ ────┤ ────┤
 │ ┌─── B │ ┌─── A │ ┌─── A
 └───┤ └───┤ └───┤
 └─── C └─── C └─── B

All three show: (A,B) are sisters, and C is outgroup. Order of tips does not matter.

Implication: Do not assume species listed next to each other are more closely related than distant ones. Only branching pattern matters.

Rooting: Determining Ancestral Direction

Unrooted tree cannot tell direction of evolution:

 A
 |
 ────┼────
 | |
 B C

Did A diverge first, or B, or C? Cannot tell.

Rooting with outgroup: Add distantly related species known to be outside your group of interest.

Outgroup (Culex) ────┬────┬─── A (Ae. albopictus USA)
 │ └─── B (Ae. albopictus Japan)
 │
 └──────── C (Ae. aegypti)

Now clear: Outgroup diverged first, then A&B lineage split from C, then A and B split.

Choosing outgroup:

Should be clearly outside ingroup (not too closely related)
But not too distant (very divergent sequences align poorly)
For Aedes phylogeny: Culex or Anopheles work well (same family, different genus)

Building Trees with IQ-TREE

Why IQ-TREE?

IQ-TREE (Nguyen et al., 2015) is the fastest and most accurate ML software for phylogenetics. Features:

Automatic model selection (tests 88 models, picks best)
Ultrafast bootstrap (1,000 replicates in minutes)
Partition models (different models for different genes/codon positions)
Tree topology tests (compare alternative hypotheses)
Free and cross-platform (Mac, Windows, Linux)

Compared to alternatives:

RAxML: Similar accuracy, but slower and no auto model selection
PhyML: Older, less accurate
MrBayes: Bayesian, much slower

Basic IQ-TREE Workflow

Command-line usage (recommended for reproducibility):

# Align sequences with MAFFT
mafft --auto mosquito_sequences.fasta > aligned.fasta

# Build ML tree with IQ-TREE
iqtree -s aligned.fasta -m MFP -bb 1000 -nt AUTO

# Parameters explained:
# -s aligned.fasta: Input alignment file
# -m MFP: Auto model selection (ModelFinder Plus)
# -bb 1000: Ultrafast bootstrap with 1000 replicates
# -nt AUTO: Use all available CPU cores

Output files:

aligned.fasta.treefile: Best ML tree (Newick format)
aligned.fasta.iqtree: Log file (models tested, tree stats, bootstrap values)
aligned.fasta.log: Progress log

Typical runtime:

50 sequences, 650 bp: 2-5 minutes
100 sequences, 650 bp: 5-15 minutes

Using Container Tools (Docker/Apptainer)

For ENTM201L: We provide containerized IQ-TREE for ease of use. Why containers?

No installation needed: Software pre-packaged with all dependencies
Reproducibility: Same environment on all computers
Isolation: Doesn't interfere with system software

Example with Docker:

docker run -v $(pwd):/data iqtree/iqtree -s /data/aligned.fasta -m MFP -bb 1000

Example with Apptainer (common on HPC clusters):

apptainer exec iqtree.sif iqtree -s aligned.fasta -m MFP -bb 1000

Visualizing Trees

Newick Format: Text Representation

Newick is a text format for trees:

((A:0.1,B:0.15)85:0.05,C:0.25);

Interpretation:

A:0.1 = tip A with branch length 0.1
(A:0.1,B:0.15) = A and B are sisters
85 = bootstrap value for node
C:0.25 = outgroup C

Human-readable but not visually intuitive. Need visualization software.

Visualization Tools

FigTree (desktop app):

Load .treefile, displays tree graphically
Adjust layout (rectangular, circular, radial)
Color branches, label nodes with bootstrap values
Export as PDF/PNG for publications

iTOL (Interactive Tree of Life, web-based):

Upload Newick tree to https://itol.embl.de
Drag-and-drop interface
Add metadata (geographic origin, date, etc.) as colored strips
Huge customization options (colors, shapes, annotations)
Export high-resolution figures

ggtree (R package):

Programmatic tree visualization (code-based)
Integrates with R data analysis workflows
Publication-quality figures with ggplot2 aesthetics
Best for: Reproducible figures, complex annotations

For ENTM201L: Use FigTree for quick visualization, iTOL for publication-quality figures.

Annotating Trees with Metadata

Example: Color Aedes albopictus samples by geographic origin. iTOL annotation file (tab-delimited):

TREE_COLORS
SEPARATOR TAB
DATA
Ae_albopictus_USA range #FF0000 Japan
Ae_albopictus_Japan range #0000FF USA

Result: USA samples red, Japan samples blue, immediately shows geographic clustering.

Real-World Application: Tracking Aedes albopictus Invasion

Background: A Global Invader

Aedes albopictus (Asian tiger mosquito):

Native range: Southeast Asia (Japan, China, Korea, Vietnam)
Invasive range: Americas (1980s), Europe (1990s), Africa (2000s)
Vector competence: Dengue, Zika, chikungunya
Invasion pathway: Used tire trade (larvae in water-filled tires)

Phylogenetic Approach

Research question: Where did USA populations originate? Methodology:

1. Sample collection: Collect Ae. albopictus from 50 sites across USA (California, Texas, Florida, New York)

2. DNA extraction + sequencing: COI barcoding (712 bp)

3. Reference sequences: Download 100 Ae. albopictus COI from GenBank (Asia, Europe, South America)

4. Alignment: MAFFT with 150 sequences

5. Phylogenetic tree: IQ-TREE with GTR+I+G model, 1,000 bootstrap replicates

6. Visualization: iTOL, color by geographic origin

Results (hypothetical but based on real studies):

 ┌─── USA West Coast (CA, WA)
 ┌─────┤ 99%
 │ └─── Japan (Tokyo, Osaka)
 │
 ┌─────┤ ┌─── USA East Coast (FL, NY)
 │ └─────┤ 95%
 │ 78% └─── Southern China (Guangzhou)
 ────┤
 │ ┌─── Europe (Italy, France)
 │ ┌─────┤ 97%
 └─────┤ └─── Northern China (Beijing)
 │ 88%
 └───────── South America (Brazil, Argentina)

Interpretation:

1. Multiple introductions: USA populations are not monophyletic (West Coast clusters with Japan, East Coast with Southern China)

2. Source populations identified: West Coast invasion from Japan (99% bootstrap), East Coast from Southern China (95% bootstrap)

3. Invasion routes: Likely via used tire shipments from specific Asian ports

4. No evidence of USA-Europe connection: European populations cluster with Northern China, independent from USA

Public health implications:

Monitor tire imports from these source regions
Expect genetic differences in USA populations (West vs. East Coast)
Different insecticide resistance profiles may have been introduced

Case Study 2: Cryptic Species in Culex pipiens Complex

Culex pipiens complex includes multiple morphologically identical forms with different biology:

Cx. p. pipiens (urban form): Breeds in polluted water, bird-biting
Cx. p. molestus (underground form): Breeds in basements, mammal-biting
Cx. p. quinquefasciatus (tropical form): Subtropical, mammal-biting

COI phylogeny (simplified):

 ┌─── Cx. p. pipiens (Europe, outdoor)
 ┌─────┤ 65%
 │ └─── Cx. p. molestus (Europe, basements)
 ────┤
 │ ┌─── Cx. p. quinquefasciatus (USA, tropics)
 └─────┤ 85%
 78% └─── Cx. p. pipiens (USA, outdoor)

Problem: Bootstrap values low (<80%), forms not clearly separated by COI. Solution: Use additional markers:

Microsatellites: Show clear genetic structure (forms are distinct)
ACE2 gene: Diagnostic SNPs distinguish forms
Behavioral assays: Mating incompatibility confirms separate species

Lesson: COI alone is insufficient for some species complexes. Phylogenetics guides further investigation.

Literature and Software Resources

Key Papers on Phylogenetic Methods

1. Phylogenetic Inference:

- Felsenstein, J. (1981). Evolutionary trees from DNA sequences: A maximum likelihood approach. Journal of Molecular Evolution 17(6): 368-376. https://doi.org/10.1007/BF01734359

- Yang, Z., & Rannala, B. (2012). Molecular phylogenetics: Principles and practice. Nature Reviews Genetics 13(5): 303-314. https://doi.org/10.1038/nrg3186

2. IQ-TREE Software:

- Nguyen, L. T., et al. (2015). IQ-TREE: A fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Molecular Biology and Evolution 32(1): 268-274. https://doi.org/10.1093/molbev/msu300

- Kalyaanamoorthy, S., et al. (2017). ModelFinder: Fast model selection for accurate phylogenetic estimates. Nature Methods 14: 587-589. https://doi.org/10.1038/nmeth.4285

3. Bootstrap Methods:

- Felsenstein, J. (1985). Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39(4): 783-791. https://doi.org/10.2307/2408678

- Hoang, D. T., et al. (2018). UFBoot2: Improving the ultrafast bootstrap approximation. Molecular Biology and Evolution 35(2): 518-522. https://doi.org/10.1093/molbev/msx281

Multiple Sequence Alignment

4. MAFFT:

- Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence alignment software version 7: Improvements in performance and usability. Molecular Biology and Evolution 30(4): 772-780. https://doi.org/10.1093/molbev/mst010

5. Alignment Quality:

- Sievers, F., & Higgins, D. G. (2018). Clustal Omega for making accurate alignments of many protein sequences. Protein Science 27(1): 135-145. https://doi.org/10.1002/pro.3290

Mosquito Phylogenetics Applications

6. DNA Barcoding and Phylogeography:

- Kamgang, B., et al. (2011). Geographic and ecological distribution of the dengue and chikungunya virus vectors Aedes aegypti and Aedes albopictus in three major Cameroon towns. Medical and Veterinary Entomology 25(2): 132-141.

- Paupy, C., et al. (2009). Comparative role of Aedes albopictus and Aedes aegypti in the emergence of Dengue and Chikungunya in Central Africa. Vector-Borne and Zoonotic Diseases 9(6): 585-595.

7. Aedes albopictus Invasion Genetics:

- Goubert, C., et al. (2016). Population genetics of Aedes albopictus invading populations. PLoS ONE 11(1): e0147673. https://doi.org/10.1371/journal.pone.0147673

- Kotsakiozi, P., et al. (2017). Population genomics of the Asian tiger mosquito, Aedes albopictus: Insights into the recent worldwide invasion. Ecology and Evolution 7(23): 10143-10157.

8. Cryptic Species and COI Barcoding Limitations:

- Cywinska, A., et al. (2006). Identifying Canadian mosquito species through DNA barcodes. Medical and Veterinary Entomology 20(4): 413-424.

- Ratnasingham, S., & Hebert, P. D. N. (2013). A DNA-based registry for all animal species: The Barcode Index Number (BIN) system. PLoS ONE 8(7): e66213.

Visualization Tools

9. FigTree: http://tree.bio.ed.ac.uk/software/figtree/

10. iTOL: Letunic, I., & Bork, P. (2021). Interactive Tree Of Life (iTOL) v5: An online tool for phylogenetic tree display and annotation. Nucleic Acids Research 49(W1): W293-W296. https://doi.org/10.1093/nar/gkab301

11. ggtree: Yu, G., et al. (2017). ggtree: An R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods in Ecology and Evolution 8(1): 28-36. https://doi.org/10.1111/2041-210X.12628

Key Takeaways

Phylogenetic Trees Reveal Evolutionary History

DNA sequences contain information about ancestry. Trees visualize this, showing:

Who is related to whom (topology)
How much divergence occurred (branch lengths)
Confidence in relationships (bootstrap values)

Multiple Sequence Alignment is Critical

Alignment ensures positional homology. Poor alignment = incorrect tree. Always:

Use robust software (MAFFT)
Visually inspect alignment
Check reading frame for protein-coding genes (COI)
Trim ambiguous ends

Maximum Likelihood is the Gold Standard

ML finds the tree that maximizes P(data | tree, model). IQ-TREE makes this:

Fast (ultrafast bootstrap)
Accurate (automatic model selection)
Easy (one command line)

Bootstrap Values Quantify Confidence

Bootstrap ≥95% = strong support. Lower values mean uncertainty. Never interpret trees without bootstrap values - topology alone is insufficient.

Outgroup Rooting Shows Direction

Unrooted trees show relationships but not ancestry. Outgroup (distantly related species) roots the tree, revealing which lineages are ancestral vs. derived.

Real-World Impact

Phylogenetics informs:

Public health: Track invasive disease vectors (Aedes albopictus)
Conservation: Identify evolutionarily distinct units for protection
Forensics: Species ID from trace evidence
Agriculture: Understand pest evolution and spread

Connection to Lab Activities

In lab, you will:

1. Collect sequences:

- Your trimmed COI sequences from Sanger sequencing

- Classmates' sequences (different Aedes samples)

- Reference sequences from GenBank (outgroups, known species)

2. Multiple sequence alignment:

- Use MAFFT (via web server or command line)

- Export aligned FASTA file

3. Build phylogenetic tree:

- IQ-TREE with automatic model selection (-m MFP)

- Ultrafast bootstrap (-bb 1000)

- Examine .iqtree file (best model, log-likelihood)

4. Visualize tree:

- Load .treefile into FigTree

- Display bootstrap values

- Root with outgroup (Culex or Anopheles)

- Color tips by species or geographic origin

5. Interpret results:

- Are your samples monophyletic (cluster together)?

- What is % divergence within species? Between species?

- Do bootstrap values support major clades?

- Can you infer invasion routes or population structure?

6. Report findings:

- Include tree figure in lab report

- Discuss bootstrap support for key nodes

- Compare your results to published phylogenies

Remember: Phylogenetics is the capstone of your molecular workflow. It transforms individual DNA sequences into evolutionary insights, connecting your lab work to global scientific questions about mosquito evolution, invasion biology, and disease transmission.

Document prepared for ENTM201L - General Entomology Laboratory UC Riverside, Department of Entomology Fall 2025

Listen to Theory Module