MLHeatmap: an interactive application for transcriptomic marker-panel discovery
Eun Young Lee, Jihye Park, Seoyeon Youn, Kyungho Choi, Sungryul Yu, Keunsoo Kang · 2026
Read MLHeatmap: an interactive application for transcriptomic marker-panel discovery by Eun Young Lee, Jihye Park, Seoyeon Youn, Kyungho Choi, Sungryul Yu, Keunsoo Kang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13472729.txt.
Description
Lee and colleagues present MLHeatmap, a browser-based program that finds small panels of marker genes from gene activity data, in 2026 bioinformatics research. Biomarker studies usually juggle separate tools for cleaning data, classification, gene ranking, and charting. MLHeatmap folds those steps into one workflow that runs from a simple count table plus group labels. It tests several classification algorithms and panel-selection methods side by side before drawing interactive heatmaps for exploration.
As a test case, the team classified colorectal cancer subtypes using 511 primary tumors from The Cancer Genome Atlas. A random forest model with forward selection reached 89.8 percent accuracy on held-out folds and produced a compact panel of just 13 genes. That panel kept a held-out accuracy score, measured as area under the curve, of 0.960 in the same data. On two independent cohorts of 500 and 159 tumors, accuracy reached 74.8 percent and 69.18 percent respectively in external testing.
Catalog record
| CREATOR | Eun Young Lee, Jihye Park, Seoyeon Youn, Kyungho Choi, Sungryul Yu, Keunsoo Kang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |