Abstract 7565: Tumor classification and deconvolution in liquid biopsy using enriched methylation sequencing
Bibliographic record
Abstract
Abstract Background: Cell-free DNA (cfDNA) detected in proximal body fluids has demonstrated potential for cancer detection using minimally invasive methodology. Our past work showed that tumor cfDNA is present in the cerebrospinal fluid (CSF) and other body fluids of patients with inconclusive standard of care testing. However, past work measuring copy number aberrations or somatic mutations was limited in cancer classification. To facilitate reliable classification, even at low tumor fractions and with fragmented DNA, we developed XR-methylSeq, a methylation sequencing platform to enrich for cell type-specific markers. Methods: We benchmarked XR-methylSeq with the K562 cell line and correlated the methylation values with gold standard measurement − whole genome bisulfite sequencing (WGBS). Methylation classifiers were applied for at least 22 cytology-positive body fluids, incorporating methylation array data from public references. T-distributed stochastic neighbor embedding (t-SNE) analysis was used for visualization in R. Deconvolution of cell type fractions for at least 29 (seven cytology-negative) body fluids and plasma samples was conducted using wgbstools. Cell type-specific markers were identified from a human DNA methylation atlas. Results: Benchmarks: XR-methylSeq has a 5-fold enrichment of the cell type-specific markers compared with WGBS. XR-methylSeq at 20 ng input highly correlates with WGBS at 2 μg (Pearson’s r = 0.97). Body Fluids: Thirteen cytology-positive CSF samples had copy number aberrations, 77% of them had concordant tumor classification, while the remaining 23% clustered with low tumor fraction samples. All nine lung primaries, including a low tumor fraction case that did not originally classify, showed a consistent cell-of-origin through deconvolution, as indicated by increased contributions from lung alveolar epithelial cells. Among six other body fluids, three exhibited the highest fractions aligning with the clinically identified cancer cell-of-origin. Additionally, the plasma cfDNA of a patient with acute liver injury had a higher fraction of hepatocyte signatures (22%) than the healthy control (8%). Conclusions: This research highlights the potential of XR-methylSeq as an enriched methylation profiling method useful for liquid biopsy applications. Citation Format: Jingru Yu, Lauren S. Ahmann, Yvette Y. Yao, Angus Toland, Alicia Snowden, Chandler Ho, Benjamin Pinsky, Hannes Vogel, Ruben Y. Luo, Linlin Wang, Brooke Howitt, Brittany Holmes, Alarice C. Lowe, Wei Gu. Tumor classification and deconvolution in liquid biopsy using enriched methylation sequencing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7565.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".