Evaluating targeted long-read sequencing as a diagnostic tool for burkitt lymphoma
Bibliographic record
Abstract
Abstract Introduction: While sporadic in developed countries, Burkitt lymphoma (BL) is endemic in resource-limited regions. Given the one-year survival rates of 61% (Mezger, The Oncologist, 2023), BL patients in sub-Saharan Africa could benefit from improved diagnostic methods, particularly those that facilitate the evaluation of new targeted agents. Recent work from our group has found that BL can now be divided into molecular subgroups based on mutations (Thomas, Blood, 2023) and into epitypes using DNA methylation patterns (Thomas, BCD, 2025). Nanopore long-read sequencing is becoming a robust and efficient approach to rapidly profile patient DNA samples for structural variants (SVs), simple somatic mutations (SSMs) and base modifications. Using the adaptive sampling (AS) technique, sequencing can be focused on regions of interest (ROIs). This technique has shown promise in application to germline genetics (Vermuelen, Nature, 2023) but its potential application to somatic conditions remains largely unexplored. We are evaluating the utility of this approach to obtain genetic and epigenetic data from relevant genomic loci to facilitate differential diagnosis and accurate assignment of BL molecular subgroup and epitype. We aim to create a diagnostic assay using AS coupled with an automated bioinformatics pipeline that classifies and subtypes BL to inform treatment and improve outcomes in resource-limited regions. Methods: To establish ROIs, we selected driver mutations and features identified as relevant to differentiating BL from other B-cell lymphomas (Coyle, Blood Adv., 2025) as well as distinguishing epitype (Thomas, BCD, 2025) and molecular subgroup (Thomas, Blood, 2023). These include 179 lymphoma genes, 3 oncogenic SVs, 127 regions commonly affected by aberrant somatic hypermutation, the EBV genome, the IGH, IGK and IGL loci, collectively 31.5 megabases. Initially, we simulated AS data for 9 BL genomes by extracting reads corresponding to ROIs with samtools from whole genome sequencing data (PromethION). To generate real AS data, we prepared 32 µL of DNA library using the Raji BL cell line. Library preparation with 1.6 µg of gDNA was performed according to the Native Barcoding Kit 24 V14 (SQK-NBD114.24) protocol. The library was sequenced on a single PromethION flow cell (R10.4.1 chemistry) for 25 hours in enrichment mode using our ROI panel. Enrichment, coverage and relevant AS metrics were assessed using samtools, bedtools, Mosdepth, custom R scripts and MinKNOW output files. To evaluate epitype classification, CpG methylation status was inferred from simulated AS data with Modkit, analyzed with custom R scripts and compared to ground truth labels. To explore 5hmC modification and hypomethylation patterns, TET2 expression was determined from RNAseq data (n=126) and analyzed using R scripts. To examine mutation differences between epitypes, SLMS-3 (Thomas, Blood, 2023) variant calls within ROIs from Illumina data for the same 9 BL genomes were used; these were normalized and compared using custom R scripts. Results: The top 6,000 5mC positions with the highest variance were able to accurately recapitulate patient epitypes. These positions had an average CpG methylation level of 49% in HyperBL and 36% in HypoBL. HyperBL samples showed more mutations than HypoBL, consistent with the ground truth. TET2 expression was significantly greater in HypoBL, suggesting more demethylation and higher transient 5hmC modification during demethylation. Regarding AS metrics, 4.7% of fragments present were basecalled. A total of 1.7 million fragments were sequenced and 23.2% overlapped a target ROI. Mean read lengths for on and off-target fragments were 7,065 and 690 bases, respectively, and the range of quality scores for these reads was comparable. Average coverage across ROIs was 67.9, with an enrichment factor of 79.3. Conclusions: Preliminary analysis of real and simulated BL AS data suggests that a limited feature set can support successful subgroup classification concordant with Illumina ground truths. Evaluation of ONT-focused tools to resolve SSMs and SVs is ongoing. Our preliminary results suggest AS has the potential to serve as a portable, high-throughput diagnostic assay to inform treatment and improve outcomes for endemic BL patients in resource-limited regions.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".