How to Analyze Nucleotide Diversity and SNPs with DnaSPAnalyzing nucleotide diversity and single nucleotide polymorphisms (SNPs) is a core task in population genetics and molecular evolution. DnaSP (DNA Sequence Polymorphism) is a widely used software package designed specifically for such analyses. This article provides a step-by-step guide on preparing data, running common DnaSP analyses, interpreting results, and integrating outputs into downstream workflows. It assumes basic familiarity with sequence alignment, population genetics concepts, and handling fasta/sequence files.
Overview of DnaSP and key concepts
DnaSP is a Windows-compatible application (can run on macOS/Linux via Wine) that analyzes DNA sequence polymorphism, linkage disequilibrium, recombination, and population genetics statistics from aligned nucleotide sequence data. Key measures you’ll commonly compute:
- Nucleotide diversity (π): average number of nucleotide differences per site between two randomly chosen sequences in the sample.
- Watterson’s theta (θw): an estimate of the population mutation rate based on the number of segregating sites (S).
- Tajima’s D: neutrality test comparing π and θw to infer demographic events or selection.
- Number of segregating sites (S): sites that are polymorphic in the sample.
- SNP frequency spectrum (site frequency spectrum, SFS): distribution of allele frequencies across segregating sites.
- Linkage disequilibrium (LD) and recombination estimates.
Preparing your data
- Sequence collection and alignment
- Gather homologous DNA sequences from your samples (e.g., mitochondrial genes, nuclear loci, whole-gene sequences). Ensure sequences cover the same region.
- Align sequences with tools like MAFFT, MUSCLE, or Clustal Omega. Inspect alignments manually in an editor (AliView, Geneious, MEGA) to check for misalignments, frame shifts, or sequencing errors.
- Trim ends to remove regions with excessive missing data or gaps so all sequences occupy the same coordinate span.
- File formats and input for DnaSP
- DnaSP accepts several input formats: FASTA, NEXUS, PHYLIP, and its own .arp/arp-like formats. Save your aligned sequences in one of these formats. For multi-locus datasets, consider separate files per locus or concatenate with partition information.
- Ensure sequence names are unique and short (DnaSP may truncate long labels).
- Missing data can be encoded as ‘N’ or ‘-’. Excessive missing data reduces the number of comparable sites and can bias statistics.
- Defining populations and groups
- DnaSP allows you to assign sequences to different populations or groups. Prepare a simple text file listing sequence names by population, or use the built-in interface to define groups. Clear, biologically meaningful grouping (by sampling location, phenotype, time point) improves the interpretability of comparisons.
Running nucleotide diversity analyses in DnaSP
- Loading data
- Open DnaSP and load your alignment file via File > Open. Confirm the displayed alignment matches expectations (sequence order, length, base calls).
- Basic polymorphism summary
- Navigate to DNA Polymorphism > DNA Polymorphism (or similar menu depending on DnaSP version).
- Choose the population or “All sequences” to analyze.
- Output will include: number of sequences (n), sequence length (L), number of segregating sites (S), nucleotide diversity (π), θw per site, and haplotype diversity (Hd).
- Sliding-window analysis of nucleotide diversity
- Use the Sliding Window function (DNA Polymorphism > Sliding-Window) to visualize local variation in π across the sequence.
- Set window size and step size appropriate for your sequence length (e.g., window 100 bp, step 25 bp for a 1,000 bp gene). Larger windows smooth noise; smaller windows detect fine-scale variation.
- Export plots or tabular results for inclusion in reports.
- Estimating confidence intervals and statistical significance
- DnaSP can compute standard errors for π and θ estimates via coalescent simulations or analytical formulas. Use the “Simulations” or “Coalescent” modules to generate null distributions for neutrality tests.
- For Tajima’s D and other tests, check p-values provided by the program—these are often derived from simulation under standard neutral model assumptions.
SNP discovery and site frequency spectrum
- Identifying SNPs
- DnaSP lists polymorphic sites and categorizes them by type (synonymous/nonsynonymous if coding sequences and codon positions provided).
- Export a table of SNP positions, alleles, and counts. This can be used for downstream analyses (e.g., genotype-phenotype association, primer design).
- Site Frequency Spectrum (SFS)
- Use the Frequency Spectrum functions (DNA Polymorphism > Frequency Spectrum) to compute folded or unfolded SFS depending on whether you have an outgroup to polarize alleles.
- The unfolded SFS requires an outgroup sequence to determine ancestral vs derived states. Without an outgroup, use folded SFS which groups minor allele counts.
- Visualize SFS to detect deviations from neutrality (e.g., excess of rare alleles suggests population expansion or purifying selection).
- SNP filtering considerations
- Exclude sites with too much missing data or ambiguous bases.
- For coding sequences, consider filtering by functional effect (synonymous vs nonsynonymous).
- When combining loci, normalize for locus length and sample size or analyze loci separately.
Tests of neutrality and demographic inference
- Tajima’s D
- Computed from π and θw. Negative values indicate excess rare alleles (possible expansion/purifying selection); positive values suggest balancing selection or population structure.
- Use DnaSP’s significance testing (coalescent simulations) to get p-values.
- Fu and Li’s tests, Fay and Wu’s H, and others
- DnaSP implements several neutrality tests. Choose tests appropriate to your data and whether you have an outgroup. Each test emphasizes different frequency classes and can help distinguish selection vs demography.
- Multi-locus comparisons and combining evidence
- Compare statistics across loci. Consistent signals across independent loci lend weight to demographic explanation; locus-specific signals point to selection.
- Consider complementary demographic inference tools (e.g., fastsimcoal, dadi) for more detailed modeling using SFS outputs.
Linkage disequilibrium, recombination, and haplotype analysis
- Linkage disequilibrium (LD)
- Use DnaSP’s LD module to compute pairwise LD statistics (e.g., D’, r^2) among polymorphic sites.
- Inspect LD decay with physical distance — rapid decay suggests frequent recombination or large effective population size.
- Recombination estimates
- DnaSP provides estimates of recombination parameters (e.g., Rm, minimum number of recombination events; ρ estimates via coalescent approaches).
- Recombination can bias neutrality tests; if recombination is high, interpret single-locus neutrality tests cautiously.
- Haplotype networks and genealogy
- Export haplotype data from DnaSP for network construction in tools like PopART or median-joining networks.
- Haplotype diversity (Hd) and network shape help visualize relationships and potential geographic/temporal structure.
Practical tips, common pitfalls, and troubleshooting
- Alignment quality is paramount: misalignments create false SNPs. Realign suspicious regions and remove poorly aligned sequences.
- Sample size matters: small n increases variance in π and SFS; report confidence intervals.
- Missing data: excessive Ns reduce effective site count—consider removing sequences or sites with too many gaps.
- Multiple testing: when running many neutrality tests across loci or windows, correct p-values (e.g., FDR) to avoid false positives.
- Version differences: DnaSP versions may differ in menu names and features; consult the program’s help for version-specific guidance.
- Reproducibility: document parameters (window sizes, filters, population definitions) and export raw tables so analyses can be re-run.
Exporting results and downstream analyses
- Export summary tables (π, θ, S, Tajima’s D), SNP lists, sliding-window outputs, and LD matrices as text or CSV for integration with R, Python, or visualization tools.
- Use R packages (ape, pegas, adegenet) or Python (scikit-allel, pyranges) to further analyze SFS, perform demographic inference, or visualize SNP distributions.
- For publication, include methods: alignment tool and parameters, DnaSP version, sequence length used, population definitions, window/step sizes, and any filters applied.
Example workflow (concise)
- Align sequences with MAFFT; trim to equal length.
- Load alignment into DnaSP; define populations.
- Run DNA Polymorphism summary for each population.
- Perform sliding-window π and export plots.
- Compute SFS (folded/unfolded) and run Tajima’s D with coalescent simulations for p-values.
- Identify SNPs and export positions; compute LD matrix and Rm.
- Export haplotypes for network visualization and feed SFS into dadi/fastsimcoal for demographic modeling.
Final notes
DnaSP remains a powerful, user-friendly tool for standard population genetics analyses focused on nucleotide diversity and SNP characterization. Careful data preparation, appropriate selection of tests, and integration with complementary tools will yield robust insights into population structure, selection, and demographic history.
Leave a Reply