Claude Skills by HolobiomicsLab

Use when when training a dual-encoder architecture (bi-encoder + cross-encoder) on NMR spectral data where independent encoding and joint pair processing produce competing or imbalanced gradient signals.

Use when you have raw mzXML LC/MS files from replicated metabolomics experiments (e.g., 12 samples across labeled/unlabeled conditions) and need to extract, align, and group peaks before downstream feature filtering (e.g., fold-change or isotope enrichment analysis).

Use when a task needs a skill from ASB Metabolomics — NMR — search this unit's 276 evidence-grounded skills, then apply and optionally ground the one that fits.

Use when running the Workflow4Metabolomics LC-MS pipeline end to end