Why Science? | sn-m3C-seq, DNA Methylation and Single-Nucleus Chromatin-Conformation Evidence

Section 1 of 36

1. Begin with one nucleus and two measurements

sn-m3C-seq links DNA methylation and chromatin-conformation information from the same nucleus. That shared origin is the central advantage: researchers need not statistically pair one methylation-only cell with another contact-only cell. Both measurements remain partial and require their own quality controls.

Contents · Next section

Section 2 of 36

2. Did you know methylation can help classify a sparse contact cell?

In the foundational human prefrontal-cortex study, methylation profiles helped identify cortical cell types, after which contact information from related cells could be aggregated. This is a principled use of one modality to organise another, provided classification confidence and specimen composition remain visible.

Contents · Previous section · Next section

Section 3 of 36

3. Separate a chemical mark from gene activity

DNA methylation records cytosine chemistry at sampled positions. Its relationship with regulatory activity depends on genomic context and cell type. A low-methylation region is not automatically active, and a contact involving that region does not by itself show that methylation caused the contact.

Contents · Previous section · Next section

Section 4 of 36

4. Keep the 3C measurement distinct

Crosslinking, digestion and proximity ligation create chimeric DNA fragments that report enriched spatial proximity. Bisulfite conversion then changes sequence information used to read methylation. The final molecule can contribute to both analyses, but each inference has different errors.

Contents · Previous section · Next section

Section 5 of 36

5. Understand bisulfite conversion

Conventional bisulfite chemistry converts most unmethylated cytosines while methylated cytosines are retained. Conversion reduces sequence complexity and complicates alignment. Incomplete conversion can imitate methylation; excessive degradation can reduce library complexity and contact recovery.

Contents · Previous section · Next section

Section 6 of 36

6. Define what joint means

Joint means that methylation and contact evidence derive from the same isolated nucleus and library design. It does not mean every locus has both measurements or that both modalities have equal coverage. A matrix should show which features are observed, missing or aggregated.

Contents · Previous section · Next section

Section 7 of 36

7. Plan cells, specimens and modalities together

Set targets for independent specimens, nuclei, methylation coverage, unique contacts and cell types. Predefine which questions use single-cell values and which require aggregation. More nuclei cannot fully compensate for one poorly represented donor or a modality that failed systematically.

Contents · Previous section · Next section

Section 8 of 36

8. Preserve tissue and nuclei before chemistry

Tissue handling, dissociation, fixation and nuclei isolation can change recovery across cell types. Record provenance, storage, processing time, nuclear morphology and DNA-content gates. A fragile neuronal population lost early will not reappear through sophisticated multi-omic analysis.

Contents · Previous section · Next section

Section 9 of 36

9. Perform in situ 3C chemistry

Crosslinked nuclei are digested and proximal DNA ends are ligated while nuclear organisation is retained. Measure digestion and ligation efficiency across specimens. Random contacts, damaged nuclei and unequal fragment recovery can alter the contact channel before methylation chemistry begins.

Contents · Previous section · Next section

Section 10 of 36

10. Sort single nuclei into wells

The published workflow uses fluorescence-activated nuclei sorting into plate wells. Keep gate definitions, index sorting, plate positions and blank wells. Doublets mix methylation profiles and contacts, sometimes creating a plausible intermediate cell that survives ordinary clustering.

Contents · Previous section · Next section

Section 11 of 36

11. Build libraries with bisulfite damage in mind

Adapters and amplification strategy must preserve enough converted DNA for sequencing. Include conversion controls and monitor fragment loss. A high conversion rate with almost no unique library complexity is not a successful joint measurement.

Contents · Previous section · Next section

Section 12 of 36

12. Track every well through amplification

Plate and cycle effects can influence read depth, duplicates and conversion. Visualise metrics by row, column and batch before biological clustering. Repeating a weak well through extra amplification can raise read count while leaving little new molecular evidence.

Contents · Previous section · Next section

Section 13 of 36

13. Sequence for methylation and contacts

Read depth planning should consider informative cytosines, unique fragments, valid contacts and cells recovered. Saturation curves reveal whether more sequencing adds loci or duplicates. Allocate coverage according to the biological question rather than requiring every cell to perform every analysis.

Contents · Previous section · Next section

Section 14 of 36

14. Use controls that test both channels

Include bisulfite-conversion controls, blank wells, known DNA where appropriate, 3C junction metrics and replicate tissue preparations. A library can pass methylation checks and fail contacts, or the reverse. Joint status should be earned by two evidence funnels.

Contents · Previous section · Next section

Section 15 of 36

15. Practise with an invented sn-m3C-seq table

These fictional values teach multi-omic review; they are not benchmarks.

Contents · Previous section · Next section

Section 16 of 36

16. Align bisulfite-converted reads explicitly

Converted DNA cannot be aligned as ordinary genomic sequence. State reference transformation, mapper, strand logic, quality thresholds and duplicate handling. Split reads spanning ligation junctions need special treatment so methylation calling and contact assignment use the correct genomic pieces.

Contents · Previous section · Next section

Section 17 of 36

17. Use TAURUS-MH or equivalent logic transparently

The foundational analysis described specialised handling for methyl-Hi-C-style reads. Whatever pipeline is used, publish its version, junction rules and intermediate files. A named tool is not a substitute for showing how reads became methylation calls and valid pairs.

Contents · Previous section · Next section

Section 18 of 36

18. Call methylation with coverage attached

For every region, report observed cytosines and uncertainty rather than a percentage alone. Sparse cells can have extreme values from very few sites. Aggregation should be restricted to appropriately classified cells and retain specimen labels.

Contents · Previous section · Next section

Section 19 of 36

19. Classify cells before aggregating contacts

Methylation features can separate broad neuronal and non-neuronal populations and finer subtypes. Use held-out markers, uncertainty and batch checks. Then show how many cells and specimens support each contact map; a confident label does not create contact coverage that was never measured.

Contents · Previous section · Next section

Section 20 of 36

20. Compare cell-type contact maps fairly

Match or downsample contacts, control genomic distance, mappability and coverage, and use biological replicates. Differences in cell number or sequencing depth can resemble stronger domains or loops. Publish distance curves and per-specimen effects alongside the aggregate heatmap.

Contents · Previous section · Next section

Section 21 of 36

21. Integrate modalities without forcing agreement

A methylation difference and contact difference at the same locus can generate a useful hypothesis. They need not point in the same direction, and neither proves the other caused it. Present concordant, discordant and unresolved loci so integration does not become confirmation bias.

Contents · Previous section · Next section

Section 22 of 36

22. Challenge incomplete conversion

Unconverted unmethylated cytosines inflate methylation estimates. Examine spike-in or non-CpG controls as appropriate and compare conversion across wells and groups. If one specimen converts less efficiently, correct the chemistry before describing a biological methylation shift.

Contents · Previous section · Next section

Section 23 of 36

23. Challenge sequence-complexity loss

Bisulfite conversion makes reads harder to map uniquely, particularly in repeats and low-complexity regions. Report mapping rates and regional mappability. Apparent absence of contacts or methylation data in a genomic compartment may reflect conversion-aware alignment limits.

Contents · Previous section · Next section

Section 24 of 36

24. Challenge cell classification circularity

If the same loci define cell types and then support the headline methylation comparison, the result can become circular. Use independent markers, cross-validation or held-out features. Contact-based conclusions should be checked under alternative reasonable cell labels.

Contents · Previous section · Next section

Section 25 of 36

25. Challenge aggregation

Combining sparse cells can reveal cell-type organisation but erase meaningful subtypes or developmental states. Display per-cell and per-specimen distributions for key features. Test whether one donor, plate or cluster boundary drives the aggregate contact difference.

Contents · Previous section · Next section

Section 26 of 36

26. Challenge causal language

Same-cell association narrows ambiguity but does not prove methylation rearranged chromatin or contacts changed methylation. Time series, perturbation and mechanism-specific experiments are needed. Use wording such as associated, coordinated or precedes only when the design supports it.

Contents · Previous section · Next section

Section 27 of 36

27. Validate with a modality-specific method

Whole-genome bisulfite sequencing or targeted methylation assays can test chemical marks; Hi-C or imaging can test conformation; RNA measurements can test expression. Choose validation that attacks the decisive inference and use independent specimens wherever the claim extends beyond one sample.

Contents · Previous section · Next section

Section 28 of 36

28. Build Primary Science observation skills

Give each paper strip a contact partner code and mark selected letters with a second colour. Learners can record both in separate columns and distinguish direct observations from inferred group patterns. The safe model shows why shared origin is valuable without assuming complete coverage or filling blanks with guesses.

Contents · Previous section · Next section

Section 29 of 36

29. Strengthen PSLE Science evidence language

A prompt can ask whether marked strips with one partner pattern are usually more or less coloured. Students must cite data, notice exceptions and avoid saying one feature caused the other. This is excellent claim-evidence-reasoning practice in an unfamiliar but manageable context.

Contents · Previous section · Next section

Section 30 of 36

30. Extend to Secondary Science chemistry and biology

Secondary learners can connect chemical conversion, DNA bases, chromosomes and cellular diversity. They can calculate percentages, compare coverage and identify confounders. The key insight is that measurement chemistry changes the sequence being read, so method and biology cannot be separated.

Contents · Previous section · Next section

Section 31 of 36

31. Use O-Level Science to evaluate the evidence

The 2026 Singapore–Cambridge O-Level Biology and Chemistry syllabuses support planning, data handling and evaluation. sn-m3C-seq provides a lively case for controls, chemical reactions, molecular inheritance, sampling and cautious interpretation across two linked evidence channels.

Contents · Previous section · Next section

Section 32 of 36

32. Approach school choices through verified opportunities

Ask official school sources how science practicals, computing, research projects, ethics and communication are supported. Programmes and admission details change, so verify them directly. A school need not run sn-m3C-seq to develop the reasoning habits that future genomics study requires.

Contents · Previous section · Next section

Section 33 of 36

33. See careers across chemistry, brains and data

This work connects neuroscience, molecular biology, analytical chemistry, sequencing technology, statistics, software engineering, pathology, ethics and data governance. Students can explore these pathways through strong fundamentals, coding, careful documentation and supervised laboratory experiences.

Contents · Previous section · Next section

Section 34 of 36

34. Use science tuition to practise two-channel reasoning

Science tuition can help students compare two tables, detect confounders, explain conversion controls and write bounded conclusions. It should complement school teaching and build transferable process skills rather than promise particular grades, admissions or biomedical careers.

Contents · Previous section · Next section

Section 35 of 36

35. Did you know one cell can hold complementary stories?

Methylation describes sampled chemical marks; contacts describe sampled spatial proximity. Reading both from one nucleus lets researchers ask more precise questions. The optimistic lesson is that integration becomes trustworthy when missing data, chemical limits and alternative explanations remain visible.

Contents · Previous section · Next section

Section 36 of 36

36. Finish with a joint-evidence checklist

Report tissue provenance, independent specimens, nuclei gates, 3C metrics, conversion controls, library complexity, alignment, methylation coverage, valid contacts, cell labels, aggregation and orthogonal validation. Release modality-specific and joint files with code. Every integrated claim should be traceable to both evidence funnels.

A defensible sn-m3C-seq study begins with two linked preregistrations: one for methylation and one for chromatin contacts. Name specimens, tissue regions, nuclei targets, plate layout, bisulfite controls, expected informative cytosines, expected unique contacts and planned cell types. Define which claims use individual nuclei and which require aggregation. A joint protocol does not justify a joint conclusion until both evidence channels meet their declared quality standards.

Tissue composition and nuclear recovery need careful accounting. Record storage, dissection, dissociation, fixation, sorting gate and blank-well behaviour. Compare recovered cell proportions with an independent reference when possible. Fragile nuclei can be lost selectively, while doublets can combine two plausible profiles. Index-sorting information and DNA content help identify those problems. Keep specimen and plate identifiers attached to every cell from acquisition through final figures.

Bisulfite conversion deserves a complete chemical audit. Report conversion control, fragment loss, mapping, unique complexity and observed cytosines per nucleus. Incomplete conversion inflates methylation, while severe degradation removes both methylation and contact information. Plot conversion against contact yield to detect joint failures. If alternative conversion chemistries are used, describe their expected base changes and validation rather than assuming the reader can transfer ordinary bisulfite expectations.

Contact reconstruction must explain split reads and ligation junctions. Save original reads, converted alignments, junction assignments, paired genomic coordinates and every filtering class. Show dangling ends, self-circles, duplicates, cis fraction, distance decay and chromosome coverage. A specialised pipeline can be excellent, but its version and parameters should be accompanied by intermediate counts. Reproducibility depends on the evidence funnel, not the software name alone.

Methylation-based classification should include uncertainty and held-out checks. Report features, dimensionality reduction, cluster resolution, marker evidence and cross-validation across specimens. Use alternative reasonable labels to repeat headline contact analyses. If one cell class appears only on one plate or donor, call it provisional. When contacts are aggregated, give the number of nuclei, donors, unique pairs and coverage per map so classification confidence is not confused with structural depth.

Integrated locus figures should display the two modalities without forcing them onto one scale. Show cytosines observed, methylation estimates and uncertainty above raw and normalised contact evidence. Include genomic distance, coverage and independent expression or accessibility data only when they add a separately measured line of evidence. Label whether the interpretation is descriptive association, temporal order or tested mechanism. Most same-cell studies establish the first and motivate the latter two.

Statistics must respect cells nested within specimens. Thousands of nuclei from one cortex do not create thousands of independent human replicates. For donor- or condition-level claims, use specimen-aware models and show leave-one-specimen-out results. Downsample contacts and methylation coverage when comparing cell types. A difference should not be driven by one group having more nuclei, better conversion or greater sequencing depth.

Data stewardship covers consent, tissue provenance, clinical context where appropriate, nuclei gates, plate maps, enzyme and conversion lots, raw reads, methylation calls, contact pairs, cell labels, exclusions, aggregates, code and software environments. Joint epigenomic and structural data can contain sensitive sequence information and require proportionate access controls. Safety includes human tissue, fixatives, sorting, hazardous conversion reagents, enzymes, heat and amplified DNA under institutional procedures.

The next experiment should target the dominant ambiguity. Repeat conversion when controls fail; collect deeper or more nuclei when one channel is too sparse; add specimens when donor effects dominate; use targeted methylation or conformation assays at a decisive locus; and perturb a regulator when causation is proposed. A good follow-up states which possible outcome supports, narrows or contradicts the model before data collection begins.

Students can use a four-box method: chemical mark observed, contact observed, association inferred and mechanism untested. Filling these boxes from a simplified figure makes an advanced multi-omic paper readable. It reinforces the difference between result and explanation, a skill that supports Primary Science observations, PSLE Science evidence, Secondary Science graphs and O-Level Science experimental evaluation.

A blind cell-quality audit can fix conversion, complexity, contact and doublet thresholds before cell-type labels are revealed. After labelling, report whether the rules remove one group or specimen disproportionately. If so, investigate whether biology altered assay performance or the filtering distorted composition. This procedure cannot eliminate every bias, but it prevents convenient thresholds from being tailored to preserve a preferred cell map.

Finally, provide a sensitivity table varying conversion cut-offs, mapping quality, cell labels, minimum contacts, methylation coverage, aggregation size and specimen inclusion. Mark which methylation-contact associations survive. A robust joint conclusion should remain visible under several reasonable pipelines and across independent specimens. When it does not, the uncertainty is itself valuable: it identifies whether better chemistry, more sampling or a targeted mechanistic experiment should come next.

Contents · Previous section · Continue to the Science Learning Hub