Tampere University Molecular Biology Master's Thesis Guide: 30 ECTS, BBT.MJS.109 and Trepo

Quick answer: what is the Molecular Biology thesis route?

Tampere University’s Molecular Biology specialisation is part of the 120 ECTS Master’s Programme in Biomedical Technology and awards Master of Science in Natural Sciences. The current programme object is BMTM.MB Molecular Biology, at least 120 ECTS; that label describes the programme-level study object and is not added on top of the 120-credit degree. The current Natural Sciences thesis route uses BBT.MJS.109 Thesis Seminar, 2 ECTS, pass/fail and BBT.OPN.002 Master´s Thesis, Natural Sciences, 30 ECTS, graded 0-5. A separate BBT.OPN.EXT.002 Research Associated with MSc Thesis, 10 ECTS, pass/fail can support laboratory, computational or clinical-data research when agreed in the thesis plan, but it is not the thesis itself.

1. Keep Molecular Biology in the Natural Sciences degree family

Molecular Biology is not the MSc (Technology) Biomedical Sciences and Engineering route. Its thesis records, seminar and degree title belong to the Natural Sciences Biomedical Technology programme. Use BBT.OPN.002 and BBT.MJS.109 as the current thesis anchors. Engineering-oriented courses or collaborators can still contribute methods, but they do not transfer the Technology thesis administration into this degree. This distinction matters for credit interpretation, maturity requirements and the final Sisu record.

2. BMTM.MB is the programme object, not extra arithmetic

Current BMTM.MB is labelled at least 120 credits and describes a programme focused on molecular mechanisms in healthy and diseased cells, molecular genetics, genomics and proteomics, protein structure, molecular interactions and organelle biology. Do not add its at-least-120-credit label to the 120 ECTS degree or to the 30 ECTS thesis. Use your own Sisu plan for exact module nesting and electives.

3. BBT.INT.802 is explicitly compulsory for Molecular Biology

Current BBT.INT.802 Orientation and Study Planning has a completion method explicitly compulsory for Cell Technology and Molecular Biology students. It combines programme sessions, laboratory training, joint introductory teaching, library/information-searching work and a personal study plan with a supervising teacher. This creates an early laboratory and information-literacy foundation, but it is not the thesis seminar: BBT.MJS.109 remains the thesis seminar.

4. BBT.MJS.109 is the 2 ECTS thesis seminar

BBT.MJS.109 Thesis Seminar, Natural Sciences is 2 ECTS and pass/fail. Students present first the thesis plan and later the results, followed by scientific discussion of background, aims, methods, interpretation, conclusions and possible future directions. Use the seminar as a quality gate: if you cannot explain biological replication, sample provenance, controls and the evidence boundary clearly, those elements are probably not sufficiently resolved in the thesis plan.

5. BBT.OPN.002 is the 30 ECTS assessed thesis

BBT.OPN.002 Master´s Thesis, Natural Sciences is 30 ECTS and graded 0-5. It supports supervised experimental or computational research and requires an accepted project plan before the project starts. The scientific report includes abstract, introduction, literature review, materials and methods, results, discussion and conclusions, and the course includes the maturity requirement. This makes wet-lab, computational and genuinely integrated molecular-biology theses possible, but the research question and evidence chain must remain explicit.

6. Keep BBT.OPN.EXT.002 separate from the thesis

BBT.OPN.EXT.002 is a separate 10 ECTS pass/fail thesis-associated research unit. It can consist of laboratory work, computational work or clinical-data acquisition and must be defined with the supervisor during initial thesis planning. It does not convert the thesis into 40 ECTS, and current public evidence does not justify calling it compulsory for every Molecular Biology student. Confirm its placement in your own Sisu plan.

7. Start with a molecular claim, not a technology list

A Molecular Biology thesis should start with the mechanism or molecular relationship being tested. “Do RNA-seq”, “study a protein” or “analyse genes” describes a technique or topic, not a research question. State the biological system, perturbation or comparison, molecular endpoint and evidence level. Then choose the assay or computational workflow that can answer that question. This prevents the project from becoming a collection of available technologies without one defensible conclusion.

8. Biological replication matters more than repeated measurements

Multiple PCR wells, sequencing reads, microscopy fields or technical replicate injections do not create independent biological evidence. Define whether the biological unit is a donor, cell preparation, tissue sample, organism, clone or experimental batch. Technical replication can improve measurement precision; biological replication is what supports generalisation across biological units. Keep the hierarchy visible in both Methods and statistics.

9. Sample provenance is part of the result

For each sample, preserve enough provenance to reconstruct source, collection or preparation context, extraction batch and downstream library/assay files. If identifiers must be pseudonymised, use a controlled key rather than removing the traceability needed for scientific quality. Sample swaps or ambiguous IDs can invalidate an otherwise sophisticated molecular analysis.

10. DNA/RNA quality must be matched to the downstream assay

Document extraction approach, input amount and relevant quality metrics when nucleic-acid quality can affect downstream results. RNA degradation, for example, can alter measured transcript profiles. Concentration or purity ratios are quality indicators, not proof that the sample is biologically valid. Define what acceptance criteria matter for the actual assay rather than collecting generic QC numbers.

11. PCR and qPCR need explicit normalization logic

For PCR/qPCR work, separate biological replicates from technical reaction replicates. Report primer/probe identity and the specificity evidence needed to interpret the assay. For qPCR, state and justify the normalization strategy, including reference genes when used. A transcript-level change remains a transcript-level observation unless additional evidence supports a protein or functional conclusion.

12. Sequencing provenance must survive the full pipeline

A sequencing thesis should connect biological sample → extracted material → library → run/file → processed dataset → statistical object → final figure. Record the reference genome/build, annotation or database versions when they influence interpretation. Sequencing depth can improve coverage but cannot replace independent biological replication.

13. RNA-seq: separate measurement, statistics and interpretation

For RNA-seq and other functional genomics analyses, report preprocessing/alignment or quantification, normalization, model/design matrix, contrasts and multiple-testing control. Differential expression is a statistical result about measured transcript abundance. Pathway enrichment, regulator inference or biological mechanism are additional interpretive layers. Label those layers so a reader can see which conclusions are directly measured and which are derived. Genome-wide genomics work should also make sample/feature filtering explicit so that a reader can reconstruct which observations entered the final model.

14. Genomic variants require reference and interpretation boundaries

Variant analysis should state the reference genome/build and annotation source. Variant detection is not the same as clinical interpretation. A candidate variant can be biologically interesting without being clinically pathogenic. If the project discusses disease relevance, separate sequence evidence, functional evidence, prior literature and any clinical classification framework rather than collapsing them into one label.

15. Functional genomics can involve several regulatory layers

Current BBT.MB.602 reflects gene regulation across prokaryotic and eukaryotic systems, epigenetic inheritance, RNAomics and systems-level biology. In a thesis, state which regulatory layer your assay actually measures: transcription, RNA processing/stability, translation, protein degradation or another mechanism. One molecular mark or expression change should not be used as universal evidence for an entire regulatory mechanism.

16. Epigenetic measurements are assay-specific evidence

If the thesis uses an epigenetic or chromatin assay, identify the molecular mark, locus or genome-wide measurement and the analytical unit. A methylation or chromatin mark may correlate with transcriptional state but does not automatically establish causation. Combine orthogonal evidence when the claim requires a mechanistic conclusion.

17. Proteomics needs identification and quantification controls

The programme explicitly includes proteomics, but a proteomics thesis still needs assay-specific reporting: sample preparation, acquisition platform, identification/quantification workflow and database/search version when relevant. If database searching is used, report false-discovery control at the level appropriate to the claim. Protein abundance, modification, interaction and functional activity are different evidence levels.

18. Protein structure: measured, modelled and predicted are different

Protein-structure work should label whether a structure is experimentally measured, computationally modelled from evidence or predicted. Do not present a high-confidence prediction as if it were an experimental structure. If a structural model motivates mutagenesis or an interaction hypothesis, let the experimental test carry the causal claim rather than the model alone.

19. Protein interaction assays need context-specific controls

An interaction detected by one assay may reflect assay geometry, abundance or indirect complex membership. State the assay and biological context, include appropriate positive/negative/competition controls where relevant, and distinguish physical binding from colocalisation or functional association. Stronger physiological claims require converging evidence.

20. Organelle localisation is not automatically direct interaction

For organelle or subcellular localisation studies, report imaging settings, segmentation/localisation rules and the biological unit used in statistics. Colocalisation can support spatial association but is not automatically proof of direct molecular binding. Multiple fields from one cell culture or tissue section remain nested within that biological sample.

21. Systems biology: networks are models of relationships

Current BBT.MJS.102 covers biological networks, gene co-expression, community detection, functional analysis and knowledge graphs. Network edges should therefore be described according to how they were constructed. A co-expression edge is statistical association, not direct molecular interaction. Record network inputs, filtering, algorithm/settings and database versions so the derived structure can be audited.

22. Bioinformatics pipelines need versioned provenance

Computational work should record software, package and database versions sufficient to reconstruct the analysis. Keep raw inputs, intermediate transformations and final tables/figures linked. If the project includes predictive modelling, separate training, validation and test data and prevent the same participant or biological source from leaking across splits through repeated samples or derived records.

23. Multiple testing is central in high-dimensional molecular data

Genomics, transcriptomics and proteomics can test thousands of features. A raw p-value threshold applied feature by feature can produce many false positives. State the multiplicity strategy, normally including an appropriate false-discovery or family-wise approach, and distinguish confirmatory hypotheses from exploratory screens. Report effect size and uncertainty where meaningful.

24. Batch effects can imitate biology

Extraction batches, library preparation, sequencing runs, reagent lots, operator/day effects and sample-processing order can create structured variation. Whenever possible, distribute biological groups across batches and record batch variables for analysis. If condition and batch are fully confounded, statistical adjustment cannot recreate the missing design information; acknowledge the limitation.

25. Human molecular and genomic data require governance

Human DNA, RNA, tissues, cells and linked molecular measurements can involve ethical approval, consent/permission and data protection responsibilities. Genomic data can remain identifiable or sensitive even after direct identifiers are removed. Establish the applicable governance and data protection route before thesis use; define storage, access and transfer controls; and keep restricted genomic or participant material outside the public manuscript where required. In genomics projects, this boundary should be designed before analysis because derived genomic datasets and variant tables may remain sensitive even when names are absent.

26. Disease claims need an evidence ladder

Separate association, molecular mechanism, model-level evidence, participant/patient evidence and clinical validation. A gene-expression change in a cell model can motivate a disease hypothesis without proving that the same mechanism controls disease in patients. Translational writing becomes stronger when it identifies what evidence exists now and what evidence would be needed next.

27. AI can assist, but it cannot own the evidence chain

Follow Tampere’s current AI guidance and agree thesis-specific use with supervisors. AI can assist with coding, language, literature orientation or debugging, but you remain responsible for scientific accuracy, citations and analysis. Do not upload protected, unpublished or identifiable molecular/genomic data to external AI systems without an approved basis. Verify generated code and molecular explanations against primary evidence.

28. Reproducibility means connecting wet lab and computation

Build an audit trail from sample identifiers and laboratory notebook entries through extraction/library files, raw instrument or sequencing data, processed datasets, scripts/notebooks and final figures. Version analysis code and record meaningful protocol deviations. A mixed wet-lab/computational thesis is most convincing when both sides of the pipeline can be reconstructed rather than only the final statistical table.

29. Write Methods while the work is happening

Molecular projects accumulate many small choices: reagent lots, thresholds, reference builds, software versions, filtering rules and exclusions. Record them during the project rather than reconstructing them from memory. Separate optimisation runs from final evaluation data and disclose analysis-rule changes made after inspecting results. This prevents a moving workflow from being presented as one fixed protocol.

30. Maturity, Turnitin, Trepo and PDF/A remain part of completion

The Natural Sciences thesis route includes the maturity requirement. Follow the current originality-checking workflow and complete Turnitin review before final submission as instructed. The assessed thesis is public, and the final archived file must follow Tampere’s current electronic archiving/PDF-A requirements. Plan restricted molecular, partner or participant information outside the public thesis.

31. Plan examiner and graduation timing separately

Thesis examination and graduation are separate processes. Tampere’s university-wide non-Technology guidance gives a normal examiner period of 21 days and up to 28 days when an additional maturity test is required. This is not the whole graduation timeline. Work backwards from graduation deadlines and reserve time for supervisor review, revisions, originality checking, final file preparation and administrative processing.

32. Final Molecular Biology checklist

Before starting: verify BMTM.MB, BBT.MJS.109, BBT.OPN.002, whether BBT.OPN.EXT.002 belongs in your Sisu plan, the accepted project plan, sample/data permissions and critical resources. During research: keep biological replication, sample provenance, batch variables, assay controls, reference/database versions and analysis code traceable. Before submission: make sure molecular observations are not inflated into clinical claims, multiple testing and uncertainty are handled, human/genomic data are protected, AI use is compliant, maturity and seminar requirements are complete, originality has been checked, and the final public PDF/A/Trepo workflow is ready.

A final evidence audit before freezing the manuscript

Before you freeze the thesis, select every central result and trace it backwards. For a qPCR figure, identify the biological samples, extraction batch, raw instrument output, normalization rule and code or spreadsheet that produced the plotted value. For a sequencing result, connect the figure to sample sheet, library/run, raw files, reference build, annotation, pipeline version, statistical object and multiple-testing output. For a proteomics result, trace peptide/protein identification and quantification to raw acquisition files and the database/search configuration. If any link is missing, either restore the traceability or narrow the result you are willing to defend.

Then perform a claim audit. Mark each major sentence as direct measurement, statistical association, computational inference, model-level interpretation, or clinical/translational implication. A result can move upward on that ladder only when the necessary evidence exists. This is especially important in Molecular Biology, where omics datasets can generate persuasive-looking pathway and disease narratives even when the thesis directly measured only transcripts, proteins or statistical relationships. The strongest discussion explains both what the data support and where the next evidence gap begins.

Finally, keep failed or negative work visible when it matters scientifically. A library-preparation batch that failed QC, a primer pair that lacked specificity, a protein interaction that disappeared in an orthogonal assay or a pathway that did not replicate can define the limits of the method and prevent future overclaiming. Negative evidence does not weaken a well-designed thesis; undisclosed selective reporting does.

Scope discipline for a 30 ECTS project

A 30 ECTS thesis does not need to answer every molecular question around a disease, protein or pathway. Prioritise one primary research question and the minimum set of orthogonal evidence needed to answer it credibly. Secondary omics, imaging or network analyses can add context, but label them exploratory when they were not part of the original confirmatory design. This keeps the report coherent and protects the project from becoming technically broad but scientifically shallow.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.

Method choices should be tied explicitly to the research question, with each assay or computational step justified by the evidence it contributes and each limitation reflected in the final claim.