LPMD Algorithm: Methylation Disorder and Azacitidine Response in AML

Xin Liu developed a Local Promoter Methylation Disorder (LPMD) algorithm to quantify epigenetic instability in acute myeloid leukemia. The dataset includes results from applying the algorithm to cohorts including GSE62298 (68 patients), TCGA-LAML (n=194), and GSE152710 (63 patients), leading to a 5-feature predictive panel. The data was last updated on March 18, 2026.

Use Cases

• Predicting azacitidine treatment response based on a 5-DMDR panel derived from multi-algorithm consensus.
• Analyzing bidirectional epigenetic disruption patterns (aberrant stabilization vs. maintenance failure) associated with DNMT3A mutations.
• Monitoring treatment dynamics by tracking LPMD changes over time, such as the significant decrease observed at 3–5 months.
• Validating methylation disorder as a biomarker across independent cohorts like TCGA-LAML and WGBS samples.

Strengths

• Algorithm validation across multiple independent cohorts, including TCGA-LAML (n=194) and a WGBS cohort (20 samples).
• The derived 5-DMDR predictive panel achieved an AUC of 0.777 with 81% sensitivity and 73% specificity.
• Longitudinal analysis includes 153 samples for treatment dynamics assessment.

Limitations

• Row count and column-level documentation are absent; field semantics must be inferred after download.
• The dataset is small (16.3 KB), indicating limited scope, likely containing summary results rather than raw patient-level data.

Provenance

Source

Xin Liu via figshare.

Collection Method

Algorithm development and analysis applied to publicly available genomic datasets (GSE62298, TCGA-LAML, GSE152710).

Freshness

Last updated 2026-03-18 05:21:24