Claude Skills by HolobiomicsLab

Use when you have a large collection of MS/MS spectra (hundreds of thousands to millions) that need to be clustered, you have already constructed nearest neighbor indexes on low-dimensional spectrum vectors (via feature hashing), and you need to compute only the relevant pairwise distances between.

Use when when you have loaded a dataset of molecular fingerprint vectors (such as biosynfoni fingerprints from a Zenodo deposit) and need to quantify how sparse the bit-representations are—that is, what fraction of bit positions are zero across the fingerprint collection.

Use when after annotating ion pairs with known adducts (using adductMatch or diffGetPeaks), use this skill to filter suspected adduct pairs by testing if their pixel-level intensity profiles are significantly correlated.

Use when after identifying candidate parent–adduct mass-difference pairs (via massdiff, histogram binning, and adductMatch), apply this skill to discriminate true molecular adducts from coincidental mass matches.

Use when you have trained predictive models (e.g., MiMeNet neural networks) on microbiome-metabolome paired data using k-fold cross-validation, held out test folds for each fold-iteration, and generated continuous abundance predictions for metabolites.

Use when you have paired predicted and observed metabolite abundance vectors from a predictive model (e.

Use when you have paired in silico and experimental measurements from multiple biological samples (e.

Use when after discovering Mass2Motifs via LDA on preprocessed MS/MS spectra, use this skill to assign putative substructure annotations by matching each motif's fragmentation pattern against a pre-indexed spectral library using learned spectral embeddings.

Use when when loading MS/MS spectra from MGF files for FIDDLE model training or evaluation, or when preparing spectrum–annotation pairs for rescore model data augmentation (TCN train/test sets).

Use when when you need to extract m/z and intensity peak values from a Spectra object backed by MsBackendMzR or similar on-disk backends; when analyzing subsets of spectra without loading all peaks into memory;

Use when you have mass spectrometry spectral data (m/z and intensity pairs, precursor m/z, MS level, and metadata) in R memory or in a file format (mzML, mzXML, CDF, MGF, MSP), and you need to organize it into a queryable, processable object before performing spectral similarity calculations.

Use when when implementing a custom MsBackend and the spectraData() method needs to return all core spectra variables (e.g., centroided, polarity, collisionEnergy) regardless of which variables were explicitly stored during backend initialization.

Use when when designing or configuring an MsBackend subclass (e.

Use when after correlation testing has validated putative parent–adduct ion pairs (e.g., via corrPairsMSI() on a massdiff object annotated with adductMatch results), use this skill to annotate and visualize the mass spectrum plot to confirm that identified pairs exhibit expected overlap—e.

Use when you have deconvolved GC-MS spectra in GNPS_GC input-compatible format and want to construct a molecular network to identify relationships between unknown compounds and perform structured chemical similarity analysis.

Use when after running spectral networking on tandem MS data and obtaining a network graph, when you need to assess which spectra cluster together, determine cluster representatives, and propagate RiPP identifications across clusters at distance 1 or 2 to enlarge the set of identified RiPPs beyond.

Use when you have computed a sparse pairwise distance matrix from MS/MS spectra (via nearest neighbor indexing) and need to partition spectra into homogeneous clusters—typically when clustering bottom-up proteomics data with the goal of grouping spectra from the same peptide sequence or when you.

Use when you have a collection of tandem mass spectrometry spectra in mzML or similar format and need to prepare them for LDA-based motif discovery.

Use when you have completed a ViMMS simulation run or processed real LC-MS/MS data and need to quantitatively assess whether one DDA controller (e.g., WeightedDEWController with exclusion) outperforms another (e.g., TopNController) in terms of spectral coverage and signal recovery.

Use when you have raw LC–MS data in vendor-proprietary or uncorrected formats (e.g., .raw, .d) and need to perform targeted peak detection, retention-time correction, or automated quality metrics on identified compounds. The input files must be converted to .

Use when when you have raw LC-MS/MS data files in mzML, mzXML, or vendor-specific formats and need to load them into a Java-based mass spectrometry analysis framework for downstream spectral analysis, fragmentation tree computation, or metabolite identification.

Use when after converting existing mass spectrometry formats (mzML, vendor formats) into mzPeak using command-line tools or when receiving mzPeak files from external sources.

Use when you have centroided mzML files from LC–MS experiments and need to perform targeted metabolomics or lipidomics analysis. Specifically, use it when you require polarity filtering, QC-sample stratification, or batch-aware peak detection—all of which depend on sampleData$type annotations (e.

Use when when you have raw LC-MS/MS DDA spectral data (positive and/or negative ionization modes) paired with sample metadata (originating taxon), and you need to detect molecular features, build a molecular network from fragmentation spectra, and annotate those features using both spectral.

Use when importing raw or public mass spectrometry spectral data in formats such as MGF, MSP, or mzML that may contain incomplete metadata (e.g., missing instrument type, precursor m/z, retention time), low-intensity noise peaks, or spectra with invalid or inconsistent metadata fields.

Use when when you have mzPeak format spectrum files and need to work with spectrum metadata, intensity/m/z arrays, or precursor information in a tabular, columnar, or vectorized computing environment.

Use when you have untargeted MS2 spectra from environmental or clinical samples that will be used for natural product identification (e.g., linking to BGCs via IOKR or other kernel-based methods), and you have access to a high-quality training library of annotated spectra with known structures.

Use when you have high-resolution tandem MS spectra (in mzML, mzXML, or MGF format) that need to be clustered or searched at scale (millions of spectra).

Use when you have a set of preprocessed MS/MS spectra (binned to 10,000 m/z bins with intensities square-root transformed) and a trained MS2DeepScore Siamese neural network model, and you need to compute 200-dimensional spectral embeddings for dimensionality reduction, clustering, or chemical space.

Use when after identifying statistically significant LC-MS features (e.

Use when when you have generated separate MemoMatrix objects from independent sample sets (e.g., sample set A and sample set B) and need to align and combine their MS2 fingerprint data into a single matrix for comparative analysis.

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 when you have a trained or untrained chemprop base model (graph convolution + readout layers) and need to extend it to predict infrared spectral properties rather than scalar molecular properties.

Use when when importing raw mass spectrometry data from instrument vendors or public repositories in one format (e.g., mzML, mzXML) and needing to export it in another format (e.

Use when you have MS2 fragmentation spectra from multiple metabolomics samples and need to compare them in a retention time-agnostic manner, especially when samples are chemically diverse, acquired with different LC methods or mass spectrometer technologies (e.

Use when you have a high-resolution LC-MS/MS spectrum or pre-computed molecular fingerprint from a small-molecule sample and need to retrieve a systematic structural classification (compound class and subclass) with confidence estimates.

Use when you have raw or unprocessed MS/MS spectral data in standard metabolomics formats (MGF, mzML, mzXML, msp, or JSON) and need to import them into a Python-based workflow for MS2 fingerprint generation, peak counting, or spectral similarity scoring.

Use when when you have metabolomics intensity data with metabolites grouped by fragmentation spectral similarity (Molecular Families or Mass2Motifs) and need to rank or score these groups by their differential activity across experimental conditions, especially when traditional pathway databases.

Use when processing raw MS/MS spectra (in MGF, mzML, mzXML, JSON, or MSP format) prior to MS2Query library matching or MS2Deepscore embedding calculation.

Use when after submitting an LC-MS/MS fragmentation spectrum to the MSNovelist web service and receiving a JSON response containing ranked de-novo structure candidates.

Use when after MS2Deepscore has selected the top 2000 candidate spectra from a library based on spectral similarity, and you need to re-rank these candidates to surface the single match (either exact or analogue) rather than rely on raw spectral similarity alone.

Use when you have multiple mass spectral library files in different formats (MSP, MGF, MOL folders) from sources like NIST, MoNA, RIKEN, or GNPS, and need to produce a single consolidated MSP file with complete SMILES, InChIKey, and experimental retention index (RI) annotations for metabolomics or.

Use when you have one or more MS/MS query spectra (in mzML, mgf, msp, mzxml, json, or pickled matchms format) and a pre-built spectral library stored in SQLite with precomputed MS2Deepscore embeddings.

Use when you have an experimental or public MS/MS spectrum (e.g., from MassBank in msp format, or a raw centroid-mode chromatogram) and need to create a reusable library entry for a known metabolite.

Use when when you have mass spectral libraries from multiple sources (NIST, MoNA, RIKEN, GNPS) in disparate formats (MSP, MGF, MOL folder structures) or with misaligned metadata (e.

Use when you have raw GC-MS output in CSV format (with Component.RT, Base.Peak.MZ, Component.Area, Compound.Name, Match.Factor, and File.Name columns) and need to systematically rank putative identifications by match quality and exact mass agreement.

Use when when you have discovered Mass2Motifs from MS2LDA topic modeling and need to automatically annotate them by finding the most structurally similar known spectra in a reference library.

Use when when you have MS/MS fragment spectra (in .mgf format) acquired from unknown metabolite features and need to annotate them against known compounds.

Use when you have tandem mass spectrometry data (LC-MS/MS in MGF, mzXML, mzML, or mzData format) and genomic data from a target organism, and you want to identify RiPPs by matching experimental spectra against a database of predicted post-translationally modified RiPP structures derived from.

Use when you have GC-MS data preprocessed into a structured spread format and need to confirm that a set of known or suspected compounds are correctly identified in your samples.