Description

This curriculum spans the full workflow of molecular evolution analysis, comparable in scope to a multi-phase bioinformatics consultancy or a structured internal genomics capability program, covering experimental design through to reproducible reporting.

Module 1: Defining Evolutionary Research Questions and Study Design

Select appropriate phylogenetic scope (intra-species, inter-species, or pan-genomic) based on biological question and data availability.
Justify inclusion or exclusion of taxa to balance representativeness and computational tractability.
Determine whether to pursue gene tree or species tree inference based on expected levels of incomplete lineage sorting or horizontal gene transfer.
Choose between de novo sequencing and public database reuse, weighing data quality against cost and novelty constraints.
Establish criteria for outgroup selection to ensure rooting accuracy without introducing long-branch attraction artifacts.
Design sampling strategies that account for geographic, temporal, and phenotypic diversity to avoid biased evolutionary interpretations.
Define thresholds for sequence coverage and quality to ensure reliable variant calling in downstream analyses.
Document metadata standards for sequences to support reproducibility and integration across studies.

Module 2: Sequence Acquisition, Curation, and Quality Control

Implement automated pipelines to retrieve orthologous sequences from GenBank, RefSeq, or SRA using BioPython or Entrez Direct.
Apply strict filtering criteria for sequence completeness, excluding entries with large unsequenced regions or ambiguous annotations.
Identify and remove chimeric sequences using BLAST-based validation or k-mer anomaly detection.
Standardize sequence naming conventions to prevent misalignment during concatenation or batch processing.
Assess contamination risks in metagenomic or environmental samples using taxon-specific k-mer profiling.
Integrate quality scores from FASTQ files into trimming decisions using tools like Trimmomatic or Cutadapt.
Validate coding sequence (CDS) annotations through ORF prediction and comparison with reference proteomes.
Document provenance and versioning of all sequence datasets to support auditability and reanalysis.

Module 3: Multiple Sequence Alignment and Homology Assessment

Select alignment algorithm (MAFFT, Clustal Omega, or MUSCLE) based on dataset size and expected divergence.
Decide between progressive and iterative alignment methods when dealing with highly divergent sequences.
Apply masking strategies to remove poorly aligned regions using Gblocks or TrimAl without over-trimming conserved motifs.
Validate alignment accuracy by inspecting conserved domain structures using Pfam or InterPro annotations.
Assess homology at the amino acid versus nucleotide level depending on evolutionary distance and selection pressure.
Handle frame shifts and indels in coding sequences by aligning at the protein level and back-translating to nucleotides.
Integrate structural alignment data (e.g., from PDB) when available to guide homology modeling in ambiguous regions.
Quantify alignment uncertainty using posterior probability scores from probabilistic aligners like PRANK.

Module 4: Phylogenetic Tree Inference and Model Selection

Compare substitution models (GTR, HKY, etc.) using AIC or BIC scores to balance fit and overparameterization.
Determine whether to use maximum likelihood (RAxML, IQ-TREE) or Bayesian (MrBayes, BEAST) methods based on dataset size and uncertainty requirements.
Set branch support thresholds (e.g., bootstrap ≥70%, posterior probability ≥0.95) for clade interpretation.
Partition data by gene, codon position, or functional domain and test for partition heterogeneity using PartitionFinder.
Account for rate variation across sites using gamma-distributed rate categories or invariant sites models.
Monitor MCMC convergence in Bayesian analyses using ESS values and trace plots in Tracer.
Address long-branch attraction through taxon addition, model refinement, or site-heterogeneous models like CAT.
Validate tree topology robustness via jackknife resampling or posterior predictive simulations.

Module 5: Molecular Clock Analysis and Divergence Time Estimation

Select calibration points using fossil records, biogeographic events, or known sampling dates with documented uncertainty distributions.
Choose between strict and relaxed molecular clock models based on empirical rate variation across branches.
Assess clock-likeness using root-to-tip regression in TempEst before applying time-scaled models.
Integrate tip-dating in BEAST for ancient DNA datasets with known radiocarbon dates.
Define priors for substitution rates based on empirical data from related clades, avoiding overly informative assumptions.
Quantify uncertainty in node ages by analyzing 95% highest posterior density intervals.
Validate temporal signal by performing date-randomization tests to rule out spurious time correlations.
Report clock model fit statistics (e.g., marginal likelihoods) when comparing alternative evolutionary scenarios.

Module 6: Detection of Selection and Adaptive Evolution

Apply codon-based models (e.g., PAML, HyPhy) to estimate dN/dS ratios across sites, branches, or clades.
Interpret ω (dN/dS) values with caution, recognizing limitations in power for weak or episodic selection.
Differentiate between pervasive purifying selection and episodic positive selection using branch-site models.
Control for recombination by screening alignments with GARD or Phi-test before selection analysis.
Validate signals of positive selection with complementary methods such as FUBAR or MEME.
Integrate population genetic data (e.g., Tajima’s D, Fay & Wu’s H) to distinguish selection from demographic effects.
Map positively selected sites onto protein structures to assess functional plausibility.
Document multiple testing corrections (e.g., FDR) when scanning genome-wide datasets for selection signals.

Module 7: Handling Recombination and Horizontal Gene Transfer

Screen alignments for recombination breakpoints using Phi-test, GENECONV, or RDP5.
Decide whether to exclude recombinant sequences, partition them, or use recombination-aware phylogenetic models.
Apply phylogenetic incongruence tests (e.g., AU test) to quantify conflict between gene trees.
Use ClonalFrameML or Gubbins to infer recombination events and correct phylogenies in bacterial genomes.
Interpret horizontal gene transfer (HGT) candidates by assessing anomalous GC content, codon usage, or phylogenetic placement.
Validate HGT events with synteny analysis across closely related genomes.
Adjust substitution rate estimates to account for recombination-induced homoplasy.
Document recombination filters and thresholds in methods sections to ensure reproducibility.

Module 8: Integration of Phenotypic and Functional Data

Map discrete phenotypic traits (e.g., drug resistance, host specificity) onto phylogenies using ancestral state reconstruction.
Test for phylogenetic signal in continuous traits using Pagel’s λ or Blomberg’s K.
Perform phylogenetic generalized least squares (PGLS) to control for non-independence in trait evolution studies.
Correlate molecular evolutionary rates with phenotypic innovation using clade-based rate tests.
Integrate gene expression or protein abundance data to contextualize selection signals.
Use phylotranscriptomic approaches to infer evolutionary changes in regulatory networks.
Validate functional predictions from evolutionary analysis with in vitro or in vivo assays when feasible.
Link positively selected sites to known functional domains using databases like UniProt or GO.

Module 9: Data Visualization, Reproducibility, and Reporting

Design publication-ready phylogenies using ggtree or FigTree, ensuring accurate scale bars and support values.
Generate time-scaled trees with annotated traits and uncertainty intervals using BEAST output and IcyTree.
Use interactive visualization tools (e.g., Microreact) for sharing spatiotemporal evolutionary patterns.
Implement version-controlled analysis pipelines using Git and containerization (Docker/Singularity).
Archive raw data, scripts, and intermediate files in public repositories (e.g., Zenodo, Dryad) with DOIs.
Adopt workflow languages (Snakemake, Nextflow) to ensure reproducibility across computing environments.
Report model assumptions, software versions, and parameter settings in detail for auditability.
Produce supplementary materials that include alignment files, tree files, and model fit statistics for peer review.