CLANC
Overview
CLANC (Comparative Latent Neoantigen Classification, or alternatively a centroid-based nearest-centroid classifier framework) is a supervised classification approach used in multi-platform cancer genomics to assign tumor samples to predefined molecular classes. In the TCGA breast cancer ILC/IDC context, it was adapted as one of several orthogonal classifiers to determine whether mixed IDC/ILC tumors are molecularly ILC-like or IDC-like, using gene expression centroids trained on pure ILC and IDC cases.
Used by
- Applied alongside ISOpure and ElasticNet as one of three orthogonal classifiers to assign 88 mixed IDC/ILC breast tumors to ILC-like or IDC-like molecular classes in the TCGA breast cancer multi-platform study (n=817); all three methods agreed that 24/88 mixed cases were ILC-like and 64 were IDC-like, with CDH1 mutation status as the dominant feature. PMID:26451490
Notes
- Three-classifier concordance (CLANC/ISOpure/ElasticNet) required for a confident call in the TCGA ILC/IDC mixed-tumor analysis.
- CDH1 mutation status was the dominant feature driving classification; all CDH1-mutated mixed cases were classified ILC-like regardless of classifier.
Sources
This page was processed by crosslinker on 2026-05-14.