Claude Skills by HolobiomicsLab

Use when when running Mass2SMILES inference on a TensorFlow-CPU build (e.g., delser292/mass2smiles:final container) and you need to optimize inference speed by controlling CPU core allocation.

Use when when training embeddings from MS/MS spectra data where you need both discriminative power (to distinguish similar spectra) and reconstruction accuracy (to preserve peak and metadata information).

Use when apply IOKR when you have BGCs with structural predictions based on MIBiG homology (cumulative BLAST score ≥10,000) and you wish to rank hypothetical BGC–spectrum links using metabolite structure information rather than genomic or strain-based features alone.

Use when a web application receives mass spectrometry data through heterogeneous identifier formats and must automatically determine which loader (Task ID, USI, or FBMN) should process the input.

Use when before attempting to run QCxMS2 for the first time, after updating any external dependencies (xtb, CREST, molbar, orca, geodesic_interpolate), or when troubleshooting unexplained calculation failures.

Use when when setting up matchms for the first time in a new environment, after upgrading Python or conda, when switching between package managers (pip vs conda), or when distributing matchms to end users to confirm functionality across supported installation channels.

Use when after cloning the ENPKG repository and installing dependencies using uv sync or conda, before executing the workflow on metabolomics datasets.

Use when you have DIA raw mass spectrometry files from multiple instrument types (timsTOF, TripleTOF, Orbitrap) and need to build a single machine learning model to predict data quality across all platforms, or when you need to compare quality characteristics of files produced by different.

Use when when preprocessing a heterogeneous spectral library (e.g., GNPS public library) that contains spectra from multiple instrument types, and you need to partition data by a single instrument class to train or evaluate a formula-prediction model.

Use when when you have raw or semi-processed mass spectrometry data files in mixed formats (e.g., vendor-native .raw, .d, .

Use when you have a large, mixed-instrument GNPS spectral dataset and need to create an instrument-specific training set for FIDDLE or similar deep learning models. Use this skill when your configuration file specifies an instrument allowlist (e.

Use when you have a time-series of repeated QCpool (pooled quality control) injections measured at regular intervals during one or more LC-MS/MS sequences, exported from Sciex Multiquant software (v3.0.

Use when when you have obtained a raw Orbitrap mass spectrometry file and need to verify that the instrument was configured as claimed in the methods section or dataset documentation—especially before investing in peptide fragmentation analysis, spectrum library matching, or quantitative proteomics.

Use when when you need to create synthetic noisy MS/MS spectra from clean baseline spectra to validate denoising algorithms, compare denoising performance across noise levels, or generate ground-truth test datasets where the true signal and noise composition are known and controllable.

Use when you have an MS/MS peak list and need to remove electronic noise—specifically when you observe suspiciously identical intensity values repeated across multiple peaks in a single spectrum, which is characteristic of instrument-generated artifacts rather than genuine analyte signals.

Use when when working with raw or filtered MsmsSpectrum objects where peak intensities span a wide dynamic range and need to be normalized for downstream spectrum comparison, database matching, or publication-quality visualization.

Use when you have processed LC-MS data (mzML or vendor format) and need to validate that internal standards and target analytes produce peak intensities within expected operational ranges.

Use when when you have loaded aligned peak data (from a preceding molecular networking alignment task) as a structured table with peak intensity, m/z, retention time, and alignment quality metrics, and you need to reduce false positives, remove noise, or focus analysis on peaks above a.

Use when you have loaded a raw or partially processed MsmsSpectrum object and need to reduce spectral noise before annotation, matching, or visualization. Use it especially when spectra contain many weak peaks (e.

Use when your lipidomics experiment includes spiked internal lipid standards of known concentration, and you have raw signal intensity matrices from LipidSearch or LIQUID output.

Use when you have extracted mass tracks (EICs) from multiple LC-MS samples aligned into a MassGrid structure, and you need to combine their intensity vectors into a single composite intensity vector for peak detection on the aggregate signal rather than per-sample.

Use when after acquiring a PRM experiment on a Thermo Fisher Orbitrap instrument when you need to verify that the mass spectrometer's data acquisition controller executed the scheduled method with correct temporal spacing.

Use when after instantiating a specXplore dashboard session layer with a loaded session data object from disk, before conducting visual exploration of LC-MS/MS spectral data.

Use when when you have extracted ion chromatograms (XICs), ion mobilograms (IMs), or mass spectra from diaPASEF or other DIA workflows and need to visualize them interactively to inspect peak boundaries, compare MS1 vs MS2 traces, validate feature identifications, or communicate results.

Use when after structural clustering (isotopologue grouping, adduct detection, cross-assay linking) and correlation clustering of LC-MS features, when you need to inspect and communicate the topology of structural relationships—particularly when the number of features or link types is too dense for.

Use when after LC-MS data has been converted to mzML format and processed through peak detection (e.g., MS-DIAL output) to yield a feature table with internal standard identifications, retention times, m/z values, and intensity measurements across multiple samples.

Use when when you have loaded extracted ion chromatogram traces (via SqMassLoader from sqMass files) and need to render them as interactive web-based visualizations where users can pan, zoom, hover for metadata, mute individual traces, and optionally visualize peak boundaries from OpenSwath results.

Use when you have resolved USI (Unified Spectrum Identifier) spectrum data from a supported repository (GNPS, MassBank, MetaboLights, Metabolomics Workbench, ProteoXchange, or MS2LDA) and need to create a figure suitable for journal publication or supplementary materials that retains a link to the.

Use when after msFeaST pipeline execution has produced a dashboard_data.json file containing quantification, metadata, and spectral matrices, or when you need to validate that preprocessing steps correctly loaded and rendered ms/ms feature data before downstream statistical or network analysis.

Use when after automated peak detection has identified candidate peaks from LC-MS mzML files, but before exporting the final metabolite library. Use this skill when you need to: (1) optimize noise and peak-detection parameters by visualizing their effect on a representative subset of peaks;

Use when after running saturation repair or multidimensional smoothing on IM-MS data when you need to validate whether corrected peaks are reliable or whether overlapping coeluting/comobiling ions may have caused incorrect signal reconstruction.

Use when after building a SummarizedExperiment object containing LC-MS peak areas and internal standard assignments, when you need to identify study samples with anomalous Internal Standard signal (indicating syringe failure, capillary clogging, or injection system malfunction) before applying.

Use when you have processed LC-MS run data (feature table or peak detection output) containing internal standard identifications and need to monitor internal standard retention time, m/z, and intensity variation across samples as part of automated or user-defined QC checks during instrument runs.

Use when you have loaded processed LC-MS data (mzML or vendor format) and need to establish baseline instrument performance before evaluating sample analytes. Use it at the start of each LC-MS batch or run to ensure that internal standard detection passes predefined thresholds;

Use when when you have tandem mass spectra data and need to predict a binary molecular property (e.g., presence of a functional group like a sulfo group) while maintaining full interpretability of the model's decision logic.

Use when when you have ensemble predictions (e.g., from Monte-Carlo Dropout inference with N ≥ 10 forward passes per input) and need to distinguish high-confidence from uncertain predictions before downstream analysis.

Use when when you have paired MS2 spectra and BGCs with structural candidates (e.g., from MIBiG homology), and you want to rank which BGC likely produces which spectrum using a compound-class-agnostic method.

Use when when you have paired genomics (antiSMASH-detected BGCs with MIBiG homology assignments) and metabolomics data (MS2 spectra from GNPS), and you need to score BGC-spectrum links using molecular structure similarity rather than strain co-occurrence patterns.

Use when you have a preprocessed feature table (m/z, retention time, intensities) from LC-MS and need to group features into empirical compounds.

Use when after loading raw diaPASEF or DIA mass spectrometry data (mzML format) and search results (DIA-NN, OpenSwath, or equivalent) to visually inspect extracted ion chromatograms and mobilograms for selected peptide precursors.

Use when your peak table contains suspected mispicked ions—ions with similar m/z and retention time that likely represent the same metabolite split across multiple features due to preprocessing errors.

Use when after applying one or more mpactr filters (mispicked, group, cv, insource) to an mpactr object and generating a qc_summary() data.

Use when when raw LC-MS feature tables exhibit inter-sample intensity variation due to instrument sensitivity drift, sample ionization efficiency differences, or loading differences, and you need to normalize intensities to a common reference scale before downstream statistical analysis.

Use when after applying one or more mpactr filters (filter_mispicked_ions, filter_group, filter_cv, filter_insource_ions) to a peak table, when you need to quantify how many ions passed or failed each filter and summarize the overall filtering impact by status distribution.

Use when when you have imzML mass spectrometry imaging data files and need to convert raw ion image intensities into quantitative lipid abundance (pmol/mm²) using known internal standards.

Use when when you have positive- or negative-mode ion mobility spectrometry data with tunemix reference standards (known m/z, drift times, and CCS values) and need to verify that the calibration model accurately captures the relationship between drift time, reference m/z, and collision cross.

Use when you have raw IM-MS data in UIMF or Agilent MassHunter .d format acquired using multiplexed (compressed) ion mobility pulse sequences, and you need to recover conventional (non-multiplexed) IM-MS spectra with resolved mobility and mass dimensions.

Use when when processing raw mass spectrometry data files of unknown or mixed provenance, and you need to automatically route IMS inputs to their corresponding analysis pipeline.

Use when when processing raw IM-MS data (Agilent MassHunter .

Use when you have raw LC-IMS-MS data (Agilent, Thermo, Bruker, or mzML format) and need to visualize and export the spatial distribution of a specific ion species (or ion family) across both ion mobility and retention time dimensions.