Show Notes
Marchi F et al., Nature Communications - This study builds the Acute Leukemia Methylome Atlas (ALMA) from 3,314 patient methylomes and presents three models—ALMA Subtype, AML Epigenomic Risk, and a 38‑CpG signature—that classify WHO2022 subtypes and predict 5‑year survival. The authors also demonstrate a nanopore-based specimen‑to‑result workflow for combined genome and epigenome profiling. Key terms: acute myeloid leukemia, DNA methylation, epigenomics, nanopore sequencing, machine learning.
Study Highlights:
The authors harmonized DNA methylation from 3,314 leukemia patient samples (331,556 CpGs) to create ALMA and a PaCMAP‑based representation that supports supervised classifiers. ALMA Subtype accurately stratifies 27 WHO2022 diagnostic subtypes and the AML Epigenomic Risk model predicts 5‑year overall survival with validation across independent pediatric and adult cohorts. A 38‑CpG prognostic signature derived by EWAS and penalized Cox modeling predicts 5‑year OS and remains independently prognostic after multivariable adjustment. A rapid long‑read nanopore specimen‑to‑result protocol on 20 specimens showed high concordance between epigenomic signatures and genomic lesions, with limitations for rare subtypes and very low coverage.
Conclusion:
ALMA and its derived models demonstrate that DNA methylation profiling, combined with machine learning and a nanopore specimen‑to‑result workflow, can improve diagnostic subtyping and 5‑year prognostication in AML, though broader training data and prospective validation are needed for rare subtypes.
Music:
Enjoy the music based on this article at the end of the episode.
Article title:
Epigenomic diagnosis and prognosis of Acute Myeloid Leukemia
First author:
Marchi F
Journal:
Nature Communications
DOI:
10.1038/s41467-025-62005-4
Reference:
Marchi F. et al., Epigenomic diagnosis and prognosis of Acute Myeloid Leukemia. Nature Communications (2025). doi:10.1038/s41467-025-62005-4
License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) – https://creativecommons.org/licenses/by/4.0/
Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00
Official website https://basebybase.com
On PaperCast Base by Base you'll discover the latest in genomics, functional genomics, structural genomics, and proteomics.
Episode link: https://basebybase.com/episodes/alma-epigenomic-aml-episode-100
QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2025-08-08.
QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the transcript's presentation of ALMA atlas construction, subtype and risk modeling, 38-CpG signature, nanopore specimen-to-result workflow, and discussed limitations. Compared these to the canonical article text.
- transcript topics: ALMA atlas construction and data scale; PaCMAP dimensionality reduction; ALMA Subtype classification across WHO2022; AML Epigenomic Risk prognostic modeling; 38-CpG AML Signature prognostic capability; Nanopore specimen-to-result protocol and clinical validation
QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 7
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0
Metadata Audited:
- article_doi
- article_title
- article_journal
- license
Factual Items Audited:
- ALMA built from 3314 patient samples across 11 harmonized cohorts (331556 CpGs).
- ALMA Subtype classifies 27 WHO2022 subtypes plus otherwise-normal control with robust cross-cohort performance.
- AML Epigenomic Risk predicts 5-year overall survival; HR ~4.40 in discovery; independent predictor after adjusting for MRD.
- 38-CpG AML Signature predicts 5-year OS; significant HRs across discovery and validation cohorts; independent predictor.
- A 38-CpG panel provides prognostic capacity with simpler data requirements compared to 331K CpGs.
- Specimen-to-result protocol uses long-read nanopore sequencing on small volumes (≈200 µL) with ~48 h turnaround; 20 specimens from 17 patients; concordance with genomic calls obser
QC result: Pass.
