BiomeLens — Integrated Bioinformatics ServicesStart a Project

Study design, effects and uncertainty

Microbiome & ecological statistics

Choose analyses that match the biological question and how samples were collected. We connect microbial community data with treatments, habitats and environmental measurements while making assumptions and uncertainty explicit.

Who this service is for

For researchers needing a defensible analysis plan, reanalysis, interpretation or clearer statistical figures for microbiome, ecological or environmental data. Support can complement an existing sequencing workflow or form a standalone statistics engagement.

Discuss this analysis

From inputs to interpretation

Analysis workflow

The final workflow is agreed after reviewing your data. Optional modules are selected for suitability and the research question.

  1. Identify the unit of replication

    Map the sampling design, planned contrasts and sources of dependence. Distinguish biological from technical replication and establish which comparisons are supported before choosing a test.

  2. Check and explore the data

    Review sample matching, distributions, zeros, missingness and confounding. Select justified transformations or normalisation, and distinguish exploratory findings from preplanned tests.

  3. Fit design-aware analyses

    Use suitable diversity models, multivariate comparisons, regression or mixed models. PERMANOVA uses appropriate permutations and dispersion assessment; constrained ordination follows the question and data structure.

  4. Assess and communicate evidence

    Review model assumptions, sensitivity and multiple testing. Report effect sizes and uncertainty alongside p-values, distinguishing robust findings from exploratory associations.

Statistics and study design

Scope may include alpha/beta diversity, PERMANOVA, differential abundance, RDA/dbRDA or CCA where appropriate, environmental associations, regression and repeated-measure or mixed models. The number of predictors and model complexity must be compatible with sample size and design.

Biological interpretation

Association does not establish causation. Sequencing depth, compositionality, batch effects, collinearity and spatial or temporal dependence can change interpretation. We explain these issues in relation to the actual dataset rather than applying the same test list to every study.

Outputs and deliverables

  • An analysis plan, data-review notes and a clear description of sampling units and comparisons.
  • Statistical result tables, diagnostics and appropriate effect/uncertainty summaries.
  • Publication-oriented figures and methods reporting; reproducible R code or other workflows as agreed.

Scientific visualisations

Ordination with clearly defined grouping, distributions showing individual observations, effect-size and confidence-interval plots, environmental association figures and annotated heatmaps. Legends explain the points, intervals and statistical annotations.

Scientific figure preparation

What if there is no biological replication?

Descriptive or exploratory analysis may still be useful, but technical repeats do not create independent biological replication. We assess the design before proposing inferential comparisons and explain when a dataset supports pattern description rather than a treatment-effect claim.

Scope your project

Tell us about your dataset.

Share sample count, grouping, replication, metadata fields, intended comparisons and existing outputs. A sample sheet or brief study-design diagram helps with scoping.

Send a brief description first. We agree data transfer, analysis scope, deliverables and a project-specific quotation before work begins.

Start a project enquiry

BiomeLens provides bioinformatics and data analysis services. Laboratory sequencing, DNA/RNA extraction and wet-lab sample processing are not part of our services.