Experimental & analytical roadmap

From paired specimens to methylation, sIgA, fragmentomics, and validation

One collection, multiple analytical views

The core pipeline generates paired tissue and plasma evidence from 18 participants. The primary study measures methylation and serum sIgA; the advisory roadmap asks which additional, defensible cfDNA features can be recovered from the same sequencing data without a second sequencing experiment.

Paired cohort

Clinical group	Participants	Plasma specimens	Paired colonic tissue	Total specimens
High-risk ulcerative colitis	8	8	8	16
Colorectal cancer	5	5	5	10
Healthy comparators	5	5	5	10
Total	18	18	18	36

Pairing provides a within-person bridge between the accessible plasma signal and the tissue context. It does not eliminate the need to account for clinical differences, cell composition, batch effects, or the limited precision of a small exploratory cohort.

Operational pipeline

flowchart TD
  P["18 consented participants<br/>3 clinical groups"] --> B["Blood processing<br/>plasma + serum aliquots"]
  P --> T["Colonic biopsy<br/>tissue preparation"]
  B --> CF["PANAMAX 48<br/>cfDNA extraction"]
  T --> GD["PANAMAX 48<br/>genomic DNA extraction"]
  GD --> SH["UltraShear<br/>tissue-DNA fragmentation"]
  CF --> EM["NEBNext EM-seq<br/>library preparation"]
  SH --> EM
  EM --> N["Illumina NovaSeq 6000<br/>paired-end sequencing"]
  B --> EL["Serum sIgA<br/>ELISA"]
  N --> Q["FASTQ + sample QC"]
  Q --> AL["fastp → bwa-meth/GRCh38<br/>→ MethylDackel"]
  AL --> DSS["DSS beta-binomial<br/>DML/DMR analysis"]
  DSS --> IN["Paired + clinical<br/>integration"]
  EL --> IN

Why enzymatic methyl sequencing

EM-seq is an enzyme-based alternative to bisulfite conversion. TET2 and T4-BGT protect modified cytosines; APOBEC deaminates unmodified cytosines before Illumina sequencing. The protocol is designed to minimize DNA damage and support low-input material, an important property for cfDNA (Vaisvila, Ponnaluri, Sun, and others, 2021).

For this study, tissue DNA is intentionally sheared. Native plasma cfDNA should not be treated identically if fragment-length analysis is planned: preserving and documenting its original insert-size distribution is a prerequisite for interpretable fragmentomics.

sIgA assay development

The reported seven-point ELISA calibration ranges from 1.250 to 0 µg/ml and produced an \(R^2\) of 0.959 in the presented run. This supports continued assay optimization, but acceptance should also consider blank behavior, standards and controls, replicate precision, dilutional parallelism, lower and upper quantification limits, and pre-specified plate-level QC—not \(R^2\) alone.

Primary methylation analysis

The planned containerized Nextflow workflow connects:

fastp for adapter and read-quality processing;
bwa-meth alignment to GRCh38;
MethylDackel extraction of cytosine-context methylation; and
DSS in R for differential loci and regions.

DSS is appropriate to evaluate because its sequencing-data model uses beta-binomial variation and dispersion shrinkage. The final model still needs a written design formula, group contrasts, pairing/subject handling, covariates, minimum coverage rules, DMR construction criteria, and false-discovery-rate control.

Paired test ≠ complete methylome model

A paired Wilcoxon test can summarize selected methylation percentages or assay-level features. It should not replace coverage-aware DML/DMR modeling across the methylome. Likewise, chi-square or Fisher tests are suitable only when categorical cell counts and assumptions support them.

Advisory extension: fragmentomics

The same paired-end plasma libraries may support a feasibility analysis of fragment size, end motifs, and nucleosome-linked coverage—feature families grounded in cfDNA biology (Lo, Han, Jiang, and Chiu, 2021; Tsui, Jiang, and Lo, 2025):

insert-length distributions and short-fragment proportions;
mono- and di-nucleosomal peaks;
coverage patterns around transcription start sites;
fragment-end positions and end-motif frequencies;
associations between fragment length and local methylation; and
integrated methylation–fragmentation summaries.

These are candidate analyses, not guaranteed free outputs. Feasibility depends on whether plasma molecules remain unsheared, library insert sizes are faithfully retained, paired-end BAM fields pass QC, depth is adequate, conversion/library steps do not introduce unacceptable bias, and a locked analysis plan controls multiplicity. FinaleMe and FRAGMA provide computational precedents, not validated drop-in models for this EM-seq dataset (Liu, Reed, Lo, Choudhury, and others, 2024; Zhou, Kang, Jiang, and others, 2022).

Follow the sequencing and genomics pipeline step by step →

External validation and systems biology

Candidate DMRs can be compared with harmonized TCGA colon and rectal cancer methylation data in the NCI Genomic Data Commons. This is external biological triangulation—not direct validation of UC progression—because TCGA tumor cohorts, platforms, biospecimens, and clinical context differ from this study.

Gene- and region-level results can then be evaluated with KEGG and Reactome for pathways involving mucosal immunity, epithelial barrier function, inflammation, and carcinogenesis. Enrichment needs an appropriate tested-region background, a recorded database version and mapping rule, effect direction, and correction for multiple comparisons.

Decision gates

Gate	Minimum evidence	Decision
Assay readiness	Extraction, conversion, library, sequencing, and ELISA QC meet locked criteria	Proceed to full cohort or troubleshoot
Primary signal	Effect sizes and FDR-controlled methylation results are stable under sensitivity analyses	Prioritize DMRs or report uncertainty
Pair concordance	Tissue/plasma direction and within-person relationships are interpretable	Advance plasma candidates or retain tissue-only evidence
Fragmentomics feasibility	Native size preservation, BAM metrics, depth, and controls are adequate	Add structural features or stop extension
External triangulation	Candidate biology is directionally coherent in relevant public data/pathways	Nominate—not declare—validation targets
Translation	Independent longitudinal cohort demonstrates reproducibility and incremental value	Consider prospective clinical-utility work

Current endpoint

The credible near-term output is an integrated candidate epigenetic–immune signature with transparent uncertainty. A diagnostic product, patent position, or claim of clinical power would require independent validation, prospective evidence, and formal institutional review.

References

Romualdas Vaisvila, Vamsi K. C. Ponnaluri, Zhe Sun, and others. (2021). Enzymatic methyl sequencing detects DNA methylation at single-base resolution from picograms of DNA. Genome Research. https://doi.org/10.1101/gr.266551.120.

Y. M. Dennis Lo, Diana S. C. Han, Peiyong Jiang, and Rossa W. K. Chiu. (2021). Epigenetics, fragmentomics, and topology of cell-free DNA in liquid biopsies. Science. https://doi.org/10.1126/science.aaw3616.

W. H. Adrian Tsui, Peiyong Jiang, and Y. M. Dennis Lo. (2025). Cell-free DNA fragmentomics in cancer. Cancer Cell. https://doi.org/10.1016/j.ccell.2025.09.006.

Yaping Liu, Sarah C. Reed, Christopher Lo, Atish D. Choudhury, and others. (2024). FinaleMe: Predicting DNA methylation by the fragmentation patterns of plasma cell-free DNA. Nature Communications. https://doi.org/10.1038/s41467-024-47196-6.

Qing Zhou, Guannan Kang, Peiyong Jiang, and others. (2022). Epigenetic analysis of cell-free DNA by fragmentomic profiling. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2209852119.