Biology › Section J · Evolution › Topic J.9

Sequence data and the tree of life

Comparing DNA sequences is how modern phylogenies are built — and the same data revealed that Darwin's tidy single-trunked tree is too simple for a world full of gene-swapping organisms.

Sequences reshaped the tree

Darwin sketched the tree of life in 1837 as a single trunk with branches, and for most of biology's history that model held. Sequencing changed the game twice over. It gave biologists a precise way to measure relatedness — compare DNA base by base and group organisms by how much they differ — and it exposed something the classic tree never allowed: genes moving sideways between unrelated species, which forced the tree to be redrawn.

From sequence similarity to kinship

Genes that share ancestry tend to keep similar sequences, because the parts that matter are preserved by selection. A newly found human gene that closely matches a well-studied mouse gene probably does a similar job. Homologous sequences come in two kinds. Orthologs are the same gene in different species, separated when the species split — human and mouse haemoglobin. Paralogs arose when a gene was duplicated within one genome, like the several haemoglobin genes humans switch between from embryo to adult. Orthology marks speciation; paralogy marks duplication.

Aligning sequences

Before two sequences can be compared, they must be lined up. Alignment places the sequences side by side so matching positions sit opposite each other, inserting gaps where one sequence has extra bases. Scoring rewards matches and penalises mismatches and gaps, so the best alignment is the one with the highest score. Two styles cover different questions: global alignment matches whole sequences end to end, while local alignment finds the best-matching segments — the right tool when only part of two genes is shared. Real phylogenetics aligns many sequences at once, but all start from this logic.

Worked example: Computing percent identity

Compare two short sequences: ACGTTA and ACCTTA. 1. Write them in alignment, one above the other, with no gaps needed — both are six bases long. 2. Compare position by position: A = A, C = C, G versus C is a mismatch, T = T, T = T, A = A. 3. Count the matches: five out of six. 4. Percent identity is matches divided by positions, 5/6, or about 83 percent. 5. But a number alone is not a conclusion: for six bases, 83 percent could easily happen by chance. Real comparisons use statistics — chiefly the E-value — to judge whether similarity is meaningful, which is the next card's job.

Worked example: Reading BLAST output

BLAST, the Basic Local Alignment Search Tool, rapidly searches a database for sequences similar to a query and lists hits with scores and E-values. Suppose a query returns: hit A with an E-value of 1e-50 and hit B with an E-value of 0.5. 1. The E-value is the number of matches that good you would expect by chance alone. 2. Hit A: expecting one chance match that good in 1 followed by 50 zeros of searches means the match is almost certainly real kinship. 3. Hit B: half a chance match is expected by luck, so it is weak evidence. 4. Rule of thumb: the smaller the E-value, the more significant the hit. 5. Bonus from aligning many sequences: positions conserved across all species often mark functionally critical regions, guiding drug design and predicting which mutations will be harmful.

Trees and outbreaks from sequence data

Phylogenetic trees are built from aligned sequences by grouping species whose sequences differ least, using distance-based methods or statistical approaches such as maximum likelihood. Molecular trees have delivered surprises the anatomy-based ones missed — famously, that the whale's closest living relative is the hippopotamus. The same machinery runs in real time on pathogens: during the COVID-19 pandemic, millions of SARS-CoV-2 genomes were sequenced and shared, letting researchers spot new variants within weeks and trace how each spread.

Horizontal gene transfer breaks the tree

The classic tree model assumed organisms pass genes only vertically, parent to offspring, with mutation supplying the variety. Horizontal gene transfer, or HGT, violates that: genes move between unrelated species. HGT is rampant in the prokaryote domains, through three classic routes plus a newer one. Transformation takes up naked DNA from the environment; transduction moves genes inside a virus; conjugation passes DNA through a hollow tube called a pilus; and gene transfer agents — small virus-like particles — shuttle random genome segments between prokaryote species. The medical consequence is severe: bacteria spread antibiotic resistance largely through these transfers, not through their own lineages alone.

Eukaryotes borrow genes too

Eukaryotes were thought exempt — until sequence data proved otherwise. Aphids, for instance, make their own carotenoid pigments, molecules humans must get from food such as carrots and mangoes. DNA analysis shows the pigment genes came from fungi, presumably picked up as aphids consumed them for food. A carotenoid enzyme called a desaturase colours certain aphids red, and when that gene is knocked out by mutation they revert to their common green. Red aphids are more conspicuous to predators, but they also resist insecticides better, so the borrowed gene can raise fitness in the right environment. HGT is a creative force, not just noise.

Genome fusion and the nucleus

The ultimate horizontal event is genome fusion: one species taken inside another's cytoplasm, their genomes merging. The Endosymbiont Theory holds that this is how eukaryotes acquired mitochondria and chloroplasts — engulfed bacteria that kept their own DNA, which is why mitochondrial DNA resembles bacterial circular chromosomes, and their double membranes are the relic of being swallowed. Some eukaryotic genes resemble archaeal genes and others bacterial ones, which fits a fusion of an archaeal host with a bacterium. But the origin of the nucleus is still contested. Three hypotheses compete: the nucleus-first model has prokaryotes gaining a nucleus before mitochondria; the mitochondrion-first model has the energy-providing symbiosis coming first; the eukaryote-first model, most radical, says prokaryotes are simplified descendants of full eukaryotes. The debate is live — sequence data has not yet crowned a winner.

Web and ring models

If genes move sideways, is the tree dead? In 1999 W. Ford Doolittle proposed the web of life: eukaryotes arose not from one prokaryote ancestor but from a pool of many species sharing genes, so the diagram needs multiple trunks and cross-connections — closer to a multi-trunked Ficus than Darwin's oak. James Lake went further with the ring of life: all three domains descending from one pool of gene-swapping prokaryotes, drawn as a ring, which his lab argues best fits extensive DNA analyses. Other phylogeneticists remain highly sceptical of the ring. The honest summary: no model — tree, web, or ring — is complete, and each stands as a hypothesis awaiting better data.

Common mistakes to avoid

First, do not treat percent identity as proof on its own — for short or few comparisons, chance produces high scores, which is exactly what E-values exist to weigh. Second, orthologs and paralogs are not synonyms: orthologs track speciation, paralogs track duplication, and confusing them scrambles the tree. Third, HGT is not mutation — mutation changes existing genes within a lineage; HGT imports genes across lineages. Fourth, antibiotic resistance spreads between bacteria by HGT, which is why it travels so fast — no bacterium needs to invent resistance from scratch. Fifth, the ring of life is a contested hypothesis, not settled fact.

Key takeaways

Sequence comparison is the engine of modern phylogeny: align sequences, score matches against mismatches and gaps, then group organisms by least difference, checking significance with E-values — smaller means more meaningful, and conserved regions mark critical functions. Orthologs split at speciation; paralogs split at duplication. The same data overturned the pure tree: horizontal gene transfer moves genes between unrelated species through transformation, transduction, conjugation, and gene transfer agents, giving aphids fungal pigment genes. Genome fusion built the eukaryote's mitochondria and chloroplasts, and life's deep shape may be a web or ring rather than a single trunk — all models pending better evidence.

Flashcards

Flashcard

Ortholog

The same gene in different species, split by speciation

Human and mouse haemoglobin are orthologs.

Test your understanding

Question 1 of 5

Human haemoglobin and mouse haemoglobin are the same gene separated when the two species split. What are they called?

Question 2 of 5

A BLAST search returns two hits: one with an E-value of 1e-50 and one with an E-value of 0.5. Which hit is the more significant match?

Question 3 of 5

Certain aphids are red because of a carotenoid enzyme gene. Where did that gene come from, according to DNA analysis?

Question 4 of 5

Antibiotic resistance spreads rapidly between bacteria. Which process is mainly responsible?

Question 5 of 5

Why did W. Ford Doolittle propose replacing the tree of life with a web model in 1999?

← Previous topicNext topic →