Claude Skills by HolobiomicsLab

Use when you have molecular structures (SMILES or graph representations) and need to predict or analyze infrared spectral properties using message passing neural networks.

Use when you receive raw spectral data files (jcamp, RAW, or mzML) from NMR, IR, or MS instruments and need to extract peak tables, metadata, and spectral parameters before generating visualizations, performing peak annotation, or converting to standardized output formats.

Use when after submitting MS/MS data and feature tables to GNPS and receiving a molecular networking job result.

Use when when you have parsed and validated spectral data (jcamp, RAW, or mzML format) from NMR/IR/MS instruments and need to create visual representations with peak annotations for inspection, annotation, or publication.

Use when you have raw spectral data from multiple complementary spectroscopic techniques (NMR, HSQC, COSY, IR) that must be combined into joint training records for a multimodal deep learning model, and the raw intensity values or chemical shift ranges differ significantly between modalities due to.

Use when you have loaded raw INADEQUATE NMR spectrum data and need to distinguish true molecular peaks from noise and instrumental artifacts.

Use when you have MS2 product-ion spectra in open formats (.mzML or .mzXML) from public mass spectrometry datasets (e.g., from MassIVE with a valid accession) and need to identify chemical compounds by comparing fragmentation patterns against the GNPS reference spectral library.

Use when you have an NMR spectrum of a mixture sample and a library of reference spectra for individual compounds, and you need to determine which compounds are present and in what proportions.

Use when you have downloaded raw spectroscopic data files (NMR, HSQC, COSY, IR modalities) from the Zenodo repositories and need to convert them into the standardized multi-modal input format required by the MultiModalSpectralTransformer before inference or retraining.

Use when you have processed LC-MS/MS spectral data (as a .mgf file with feature identifiers) and computed pairwise ms2deepscore similarity scores, and you need to create a 2-D projection suitable for dashboard visualization or high-level pattern discovery without losing similarity structure.

Use when when you have received POST requests containing peaks data as form parameters (chemical shift, multiplicity, integration values) and need to accept, validate, and normalize those values before formatting them into a query compatible with an external NMR prediction service such as.

Use when you have a raw 1D NMR spectrum (FID or processed format) from urine, worm, or other biological samples and need to extract peak positions, intensities, and linewidths as a tabular peak list.

Use when you have raw INADEQUATE NMR spectra files and need to transition from continuous spectral data to discrete peak coordinates. Use it as the first signal-processing step before clustering peaks into networks or matching against metabolite databases.

Use when after peak picking has been completed on INADEQUATE NMR spectra and you need to group correlated peaks into compound-specific networks before matching against a metabolite database.

Use when before feeding a peak list into the NMRformer model or other transformer-based spectral assignment frameworks.

Use when you have raw spectroscopic datasets (NMR, HSQC, COSY, IR) in standardized array or DataFrame format and need to curate them for multimodal transformer training.

Use when you have a query mass spectrum (or representative metabolite spectrum from public data) and need to identify it by searching against large spectral reference databases (millions to billions of spectra).

Use when when you have identified a spectral window of interest in a 1H NMR spectrum from a complex biological sample (serum, urine, CSF, tissue, saliva, or sweat) and need to systematically retrieve all metabolites from HMDB whose reference proton NMR chemical shifts fall within that window.

Use when after extracting and optionally combining MS2 spectra from a chromatographic peak (e.g., at a known m/z value like 304.1131), you need to determine which compound(s) in a reference library match the experimental spectrum.

Use when after generating TP candidates (via in-silico prediction or library lookup) and extracting MS/MS peak lists for both parent features and TP feature candidates, use spectral similarity scoring to quantify fragmentation pattern overlap.

Use when you have preprocessed 1D ¹H and/or ¹³C NMR spectra from an unknown organic compound with ≤19 heavy atoms, and you need to recover its molecular structure (both formula and connectivity) rapidly without access to 2D NMR experiments (HSQC, HMBC, COSY) or mass spectrometry.

Use when after peak clustering, network filtering, and database matching have identified candidate metabolites and their associated peak networks.

Use when when you have raw spectroscopic measurements in heterogeneous formats (IR, Raman, UV-Vis, mass spectra, or NMR) and need to feed them into a spectrum-conditioned diffusion model for de novo molecular structure elucidation.

Use when use STOCSY when you have preprocessed 1H NMR spectral data with an unidentified peak of interest (driver signal at a specific δ ppm value) and need to determine its metabolite identity by finding correlated signals across the spectrum.

Use when you have 2D NMR spectral data (HSQC, HMBC, COSY) from multiple samples in a library or mixture and need to identify which structural scaffolds are shared across samples, prioritize samples for further analysis based on scaffold novelty or frequency, or characterize the core structural.

Use when when you have paired MS/MS spectra from unknown metabolites and a reference database of known metabolites, and you want to rank candidate structures for unknown compounds by their predicted structural similarity rather than exact spectral matching.

Use when you have experimental mass spectrometry spectra (LC–QTOF or similar format) from unknown compounds and need to predict their molecular structures.

Use when after executing a molecular structure prediction model on spectroscopic input data and obtaining predicted molecular structures in a standardized format (e.g., SMILES, MOL, SDF).

Use when after an NMR-based structure prediction model has generated predicted molecular structures (formula and connectivity) for a test set of molecules with up to 19 heavy atoms.

Use when after applying jsonschema validation to a parsed mwTab file against MS or NMR schema definitions, when you have collected validation errors, warnings, and metadata column matching results and need to communicate findings to data curators or submitters in a structured, machine-readable.

Use when you have cross-validated, filtered metabolomic NMR or MS data in a SummarizedExperiment container and need to prepare it for metabolome-wide association studies (MWAS) with epidemiological confounders.

Use when when you have a SummarizedExperiment containing metabolomic abundances and a corresponding vector of quality metrics (e.g., coefficient of variation computed across QC samples), and you need to filter to retain only features meeting a reproducibility threshold (e.g., CV ≤ 0.

Use when you have a CSV or table-format spectral peak list (with chemical shift, intensity, and metadata columns) destined for NMRformer or similar peak-to-metabolite assignment models, and you need to exclude low-quality peaks that would otherwise harm prediction accuracy.

Use when when converting mwTab-formatted metabolomics files (containing MS/NMR tabular data blocks) to JSON, or when you need to extract, manipulate, and re-serialize tabular sections from mwTab files while maintaining column structure and type information.

Use when you have nuclear magnetic resonance (NMR) peak data (1H and 13C measurements) that you need to classify using a deployed SMART 3 model, and you want to submit peaks programmatically rather than through a web UI.

Use when after completing Docker installation and container build steps for CloMet, before attempting substantive data analysis or pipeline execution.

Use when a machine learning model produces multiple ranked predictions (each with an associated confidence score) for a single input, and you need to quantify how often the correct answer appears in the top-k predictions.

Use when you have acquired or generated multi-modal spectroscopic data (integrated NMR, HSQC, COSY, IR spectra) in the model's expected input format, a pre-trained MultiModalSpectralTransformer checkpoint is available, and you need to predict molecular structures from these spectra without.

Use when you have a sequence-to-structure prediction task where: (1) inputs are token sequences or embeddings representing molecular fragments or spectral data; (2) outputs are structured molecular representations (adjacency matrices for connectivity, formula vectors);

Use when when you have CNN-encoded spectral features (¹H and/or ¹³C NMR) and a set of predicted or candidate molecular fragments, and you need to determine which fragments are present and how they connect to form a valid molecular structure.

Use when you have preprocessed 1H NMR spectral data with compound labels and need to identify multiple compounds in a flavor mixture where both local spectral patterns (handled by CNN) and long-range spectral dependencies (handled by Transformer) are diagnostic.

Use when when processing sequential spectroscopic data (1H NMR spectra) where both local chemical shift patterns and global spectral dependencies are needed for compound classification.

Use when you have a 1D 1H NMR spectrum represented as chemical shift vs. intensity values, a corresponding list of detected peak positions, and you need to assign metabolite identities to those peaks.

Use when you have aligned feature tables from LC–MS/MS, corresponding in silico annotations (from GNPS/ISDB or SIRIUS), and metadata describing sample origin. Use it to rank extracts by the proportion of sample-specific, unannotated features—a proxy for structural novelty.

Use when you have experimental LC–QTOF mass spectra from complex environmental samples (e.

Use when when a parsed mwTab file (MS or NMR experimental data) must be assessed for conformance to its corresponding JSON schema specification. Apply this skill after loading the mwTab file using the mwtab parser but before quality assurance sign-off or deposition to the Metabolomics Workbench.

Use when you have a Python application (e.g., ROIAL-NMR) with documented dependencies and version constraints, and you need to install it on a fresh machine or verify that the environment can be reconstructed without import or runtime errors.

Use when when you have both an observed NMR mixture spectrum and a candidate reconstructed spectrum (each represented as intensity distributions across chemical shift bins), and you need a scalar similarity metric to evaluate how closely the reconstruction matches the observed data.

Use when you have downloaded a web application (e.g., COLMARvista) that uses WebWorker and WebAssembly components and need to run it locally by opening index.html in a browser, rather than accessing it through a web server.

Use when you need to run a local HTML file that uses WebWorker or WebAssembly (such as COLMARvista for NMR spectra analysis) and the browser raises cross-origin or file-access policy errors preventing WebWorker initialization or WebAssembly module loading.