Claude Skills by HolobiomicsLab

Use when you have a trained PS2MS deep learning model, a set of evaluation compounds (especially novel NPS analogues), and want to understand whether prediction confidence (softmax probabilities, uncertainties) degrades gracefully or sharply as structural distance from the training set increases.

Use when you need to confirm that omitting an optional input parameter (such as secondaryAssay in buildExperiment) produces the expected mathematical result—specifically, when a default value should neutralize a transformation (e.

Use when you have instantiated a learned component (embedding layer, encoder, or transformer submodule) from a published codebase and need to verify that its forward pass produces outputs matching the original paper's implementation before integrating it into a downstream analysis pipeline.

Use when when calling filter functions (e.g., filter_mispicked_ions(), filter_group(), filter_cv()) on R6-based metabolomics data objects in the mpactr package and you need to verify whether the original object's state is preserved.

Use when after training multiple MLPNN models (via cross-validation) on paired microbiome and metabolome data when you need to extract interpretable feature importance scores from network weights to identify which microbes drive metabolite predictions and group them into co-functional modules.

Use when when you have a webservice codebase (Python, Java, etc.) with HTTP route definitions, parameter handling, and serialization logic, and you need to generate an OpenAPI 3.

Use when when a Python package provides optional support for specialized data formats or functionality (e.

Use when you are implementing ORA for metabolomics pathway analysis and must decide which metabolites constitute the statistical background against which to test your experimental detection list.

Use when when you have a Thermo Scientific .raw file from an Orbitrap instrument (e.

Use when you have a Thermo Fisher Orbitrap .raw file and need to retrieve a specific scan's spectral data (m/z and intensity arrays), validate instrument parameters (resolving power, AGC injection time), or assess signal-to-noise characteristics of fragment ions for a known precursor peptide (e.

Use when when building a multi-backend visualization library where users specify both a plot type (spectrum, chromatogram, peakmap) and a backend (matplotlib for static output, Bokeh or Plotly for interactive), and you need to avoid code duplication across backends while keeping the user-facing API.

Use when you have a flat file of structure-organism pairs (e.

Use when after organism names have been cleaned and standardized (e.g., via 1_cleaningOriginal.R and 4_cleaningTaxonomy.R) but before final integration of organism, structure, and reference data. Use it when you have a cleaned organism table (interim/tables/2_cleaned/organism/cleaned.tsv.

Use when when a project JSON document contains genome identifiers but lacks corresponding organism name annotations, and you need to link MS/MS mass spectra with genomic context for downstream biosynthetic gene cluster or chemical ecology analysis.

Use when you have a cleaned organism table (with validated, deduplicated organism names from sources like NCBI, manual curation, or previous cleaning steps) and need to annotate each organism with Open Tree of Life identifiers, taxonomic lineage, and standardized taxonomic annotations.

Use when after organism name cleaning and taxonomy verification (4_cleaningTaxonomy.R) have been completed and you have a cleaned organism table with standardized names.

Use when when developing or extending mass spectrometry data processing workflows (e.

Use when when you have run ORA on a metabolomics dataset and obtained p-values for pathway enrichment, but you need to assess whether observed significance is genuine or an artifact of incomplete metabolite coverage.

Use when when training a fresh CNN model from scratch on labeled LCMS data (e.g., MS1 peak classification in NeatMS), particularly when aiming for a specific performance target (e.g., AUC ROC > 0.9) and you need to avoid wasting compute time on redundant epochs or degrading validation performance.

Use when you need to set up a cloned or downloaded scientific Python package for local development, testing, or execution. Specifically, use it when: (1) you have a package repository with a pyproject.toml or setup.py that declares dev dependencies;

Use when a Python package has been relocated to a new repository location, reorganized to conform to new organizational standards (e.g., metabolomics-cloud conventions), or its dependencies, metadata, or CI/CD workflows have been modified.

Use when you need to verify that a Python package (or similar installable software) passes its declared integration test suite as a prerequisite to trusting its reliability in production or downstream analysis. Specifically, apply it when you observe a periodic testing CI workflow badge (e.

Use when when you need to establish a working installation of a Python package in a fresh or isolated environment, particularly when the package is available through multiple distribution channels (PyPI, Bioconda, AUR) and you want to verify that the installation is complete and functional before.

Use when releasing a new version of a Python package to public repositories, when verifying that distribution pipelines are functioning after code changes, or when troubleshooting installation failures reported by users across different platforms (Linux, macOS) or architectures (x86_64, aarch64).

Use when after creating a fresh conda environment from a pinned dependency specification (environment.yml or requirements.txt) and installing packages via conda and/or pip.

Use when when processing LC-MS metabolomics datasets with 10 or fewer samples and requiring reproducible mass track alignment across the cohort.

Use when you have a collection of preprocessed mass spectra (in mzML, mzXML, msp, MGF, or JSON format) and need to quantify similarity relationships across all pairs—for instance, to identify redundant spectra in a library, cluster related compounds, or perform spectral library searches.

Use when when you have raw mass spectrometry data (from mzML, Bruker .d, or CSV format) loaded into a Pandas DataFrame and need to ensure it has the correct column structure (m/z, retention time, intensity) before invoking pyOpenMS-Viz plotting functions like .plot(kind='spectrum'), .

Use when your input consists of multiple large MSP files (hundreds of megabytes) with associated structure folders containing hundreds of thousands of MOL or SDF files that require sequential processing steps (structure extraction, SMILES matching, RI assignment).

Use when when you have a Spectra object backed by an on-disk MS data source (e.g., MsBackendMzR reading mzML, mzXML, or CDF files) and need to process large numbers of spectra in parallel or serial chunks.

Use when designing a contrastive learning pipeline for ion images or other data modalities where you need to process multiple augmented versions of the same input through an encoder and enforce similarity between the resulting representations.

Use when when you have applied biotransformation rules to generate candidate product structures from a set of input molecules (represented as SMILES strings) and need to document which product structures were derived from which parent structures.

Use when when implementing or extending a DSL parser (lexer + recursive descent or LALR parser) that accepts user-authored query strings.

Use when you have a log2-transformed, standardized peak intensity matrix (rows = metabolite features, columns = samples) with compound annotations mapped to curated pathway databases (KEGG, Reactome, or custom metabolite sets), and you need to rank pathways by their activity level while tolerating.

Use when when preparing to run Over-representation Analysis (ORA) on a metabolomics study, after constructing the background set but before running the enrichment test.

Use when you have run a metabolomics experiment with incomplete coverage of a reference pathway database (e.g., 10–100% of database metabolites detected), and you plan to use ORA for pathway enrichment. The skill is essential if your detection method has known sensitivity limits (e.

Use when you have intensity measurements (peak features, protein intensities, or gene expression values) with compound or gene annotations (KEGG IDs, ChEBI IDs, UniProt IDs, or ENSEMBL IDs), and you need to aggregate them into biologically meaningful pathway groups for differential analysis.

Use when when preparing to run ORA on a metabolomics study: you have a list of detected metabolites from your experiment and need to determine which metabolites from the full pathway database should serve as the statistical background, and which pathways contain how many metabolites overall.

Use when you have a metabolite intensity matrix (samples × metabolites) with assigned annotations (peak IDs mapped to KEGG or ChEBI compound IDs), a metabolic pathway database, and need to rank pathways by activity level within experimental groups or comparisons.

Use when when you have completed a PALS pathway analysis on a clean metabolomics peak intensity matrix and pathway annotation set, and you need to verify that the ranked pathway discoveries are not artifacts of favorable data quality.

Use when after peak annotation when you have: (1) a peak intensity matrix (rows=peaks with KEGG/ChEBI/UniProt IDs, columns=samples) with group labels; (2) a pathway database (KEGG, Reactome, or user-defined metabolite sets); (3) a comparative experimental design (case vs. control groups).

Use when after peak detection and before any comparative analysis (e.g., diversity indices, ordination, or statistical testing) when working with direct injection FT-ICR MS data where raw peak intensities vary across samples due to instrumental factors. The task_id=task_005 example applies it to S.

Use when you have a tandem mass spectrum (MsmsSpectrum) from a known peptide and need to determine what fraction of observed peaks can be explained by expected fragment ions.

Use when after training or loading a NeatMS neural network model, before applying it to filter false positive MS1 peaks in a new dataset.

Use when apply peak-count capping when preprocessing tandem mass spectrometry (MS/MS) spectra for peptide identification or spectral library matching, particularly when working with high-resolution spectra that may retain numerous low-intensity noise peaks after intensity filtering.

Use when when you have initialized an MsBackend subclass (e.

Use when after generating simulated mzML output from ViMMS and you need to compare it against real acquisition data. Specifically, use it when you have paired real and simulated mzML files from the same sample (e.g., Beer1pos) and must evaluate whether a fragmentation strategy (e.

Use when converting raw MS/MS spectra from library files (e.g., .msp format) into structured library entries, or when annotating experimental LC–MS features against fragment databases.

Use when after scipy.signal.find_peaks has identified candidate peaks on a composite mass track segment, evaluate each peak to decide whether to retain it in the final feature table.

Use when you have raw LC-MS/MS spectra from vendor instruments (mzML, mzXML, MGF, or MSP format) with variable peak quality and intensity distributions, and you plan to perform library matching, molecular networking, or spectral similarity comparison.