Genomics, Systems Biology and Bioinformatics

Unified framework for multi-omics teaching

One framework, two pathways

• Reuse concepts and worked examples across both harmonized MSc outlines.
• Select depth and exercises to match each pathway’s learning objectives.
• Keep study design, computation and biological interpretation connected.

Use the teaching materials

1. Read the module and identify the measurement and assumptions.
2. Run the practical and predict the effect of one change.
3. Attempt exercises before opening suggested answers.
4. Explain what the result supports and what needs further evidence.

Course map · Modules · Practicals

B01 Biological questions

• Define the production trait, population and decision time.
• Separate intervention, association and prediction.
• Choose measurements that answer the question.

Discuss: Does a day14 profile qualify as a baseline predictor?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B02 Design and sampling

• Identify the independently allocated experimental unit.
• Cross treatment groups with processing batches.
• Record animal, specimen, tissue, time and handling.

Discuss: Can adjustment separate perfectly confounded treatment and batch?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B03 Databases and identifiers

• Distinguish study, sample, gene, transcript and protein records.
• Preserve species, assembly and annotation versions.
• Audit unmatched and one-to-many mappings.

Discuss: Does a duplicated annotation row create a new measurement?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B04 Reproducible data handling

• Define feature and sample dimensions.
• Align by stable identifiers; check uniqueness and missing matches.
• Preserve raw inputs and record transformations.

Discuss: Would equal dimensions detect a sample-order error?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B05 Genomic measurements

• Distinguish reads, assemblies, alignments and genotype calls.
• Specify quality encoding and the called-base denominator.
• Separate identity, homology and functional claims.

Discuss: Why can a high-identity short sequence be weak evidence?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B06 Transcriptomics

• Define biological replicates and the expression contrast.
• Separate exposure normalization, visualization and inference.
• Count-model variance and dispersion affect uncertainty.

Discuss: Do many reads replace independent animals?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B07 Epigenomics

• Record coverage, tissue and assay context.
• Distinguish within-cell changes from composition shifts.
• Regulatory association is not evidence of mediation by itself.

Discuss: Can bulk methylation change while each cell type stays constant?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B08 Proteomics

• Separate peptide identification, protein inference and abundance.
• Preserve the measurement scale and observation counts.
• Missing abundance is not a measured zero.

Discuss: What does a one-unit log2 difference represent?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B09 Metabolomics

• Distinguish analytical features from identified compounds.
• Assess blanks, standards, drift and batch.
• A standard corrects only the variation it tracks.

Discuss: Can an unresolved feature establish a named pathway?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B10 Microbiomics

• Define sampling site, time and profiling method.
• Relative read fractions do not measure total microbial load.
• Host associations can share diet, genotype and technical drivers.

Discuss: Can a fraction decrease while absolute amount stays constant?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B11 Phenomics

• Define phenotype units, time window and coverage.
• Repeated sensor rows remain nested within animals.
• Use only information available at the decision time.

Discuss: How many independent allocations are in a million sensor rows?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B12 Statistics and validation

• Distinguish conditional inference from prediction.
• Fit preprocessing and tuning within training data.
• Evaluate on independent units matching deployment.

Discuss: When does a test set become a tuning set?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B13 Association and regulation

• State counted alleles, covariates and relatedness assumptions.
• LD can make several candidates tag the same signal.
• GWAS, eQTL, TWAS and prediction answer different questions.

Discuss: Does a correlated eQTL and trait signal prove mediation?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B14 Pathways and networks

• Define the eligible background and selection null.
• State what an edge measures and how it is selected.
• Enrichment and correlation do not directly measure causal activity.

Discuss: How would changing the background alter enrichment?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B15 Multi-omics integration

• Match animal, tissue and time across layers.
• Choose an integration strategy and justified block weights.
• Evaluate missingness and shared biological/technical drivers.

Discuss: Are two correlated layers independent replications?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B16 Advanced methods

• Cell and spatial resolution do not create independent allocations.
• Keep specimens/animals separated for new-animal prediction.
• Evaluate measurement, model and mechanistic evidence separately.

Discuss: Can spatial proximity establish a molecular interaction?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B17 Research integrity

• Document provenance, exclusions and exploratory decisions.
• Share only data allowed by the governing agreements.
• Verify and disclose AI contributions appropriately.

Discuss: Does a checksum certify biological validity?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

B18 Animal-production synthesis

• Connect each assay to the defined production question.
• Compare strategies on the same independent evaluation animals.
• Report uncertainty, alternatives and justified follow-up.

Discuss: Does a predictive improvement identify an intervention target?

Use the corresponding B-numbered module for equations, worked code and suggested answers. Ask students to identify the analysis unit and evidence needed before interpreting the example.

Worked thread through the course

• Plan an animal-level diet study and matched assays.
• Verify sample identity and assay-specific quality.
• Analyse RNA counts and expression contrasts.
• Examine pathways and cross-layer relationships.
• Compare predictive strategies with training-only transformations.
• Write an interpretation or proposal at the level the evidence supports.

Complete the pathway

GSB: data resources → computation → association → networks → integration.

Multi-Omics: technologies → sampling/design → processing → integration → production applications.

Both use the same maintained module and practical sources. The map preserves original module numbering and assessment weights.