Claude Skills by HolobiomicsLab

Use when importing MS/MS spectral libraries (particularly from MoNA or GNPS) where SMILES or chemical structure identifiers are embedded in free-text or non-standard Comment fields rather than in dedicated SMILES/InChIKey fields, or when positive and negative ionization mode spectra are commingled.

Use when when you have raw feature tables exported from a tandem LC-MS/MS preprocessing tool (e.g., Progenesis QI, MS-DIAL, Bruker Metaboscape) and need to combine them with sample metadata (group assignments, replicate structure) before applying feature filtering or quality control workflows.

Use when when preparing paired MS/MS spectra for training or validation of a siamese neural network model, and you have chemical structure annotations (InChI, SMILES, or InChIKey) for each spectrum but lack pre-computed structural similarity labels.

Use when when you have a set of MS/MS spectra with ground-truth structural similarity labels (Tanimoto scores computed from molecular fingerprints) and need to choose a decision threshold for classifying spectrum pairs as 'chemically related' or 'unrelated'.

Use when you have a collection of molecular fingerprint vectors (such as biosynfoni count fingerprints) and need to measure structural similarity between all pairs of molecules.

Use when when you have centroided .mzML LC–MS runs and a target list (compound ID, theoretical m/z, expected RT, polarity) but are uncertain whether your m/z and RT windows are wide enough to capture all targets without false positives.

Use when you have centroided LC-MS data (.mzML format) and a curated list of targeted metabolites or lipids (with m/z, retention time, and polarity) that you want to quantify and quality-assess across multiple analytical runs, and you need both per-run AUC values and averaged QC metrics for each.

Use when you have centroided mzML LC–MS data, a curated list of target compounds (with theoretical m/z, expected retention time, and polarity), and you need to confirm target presence and extract quantitative metrics (area under curve, max intensity, signal-to-noise ratio, peak correlation, point.

Use when when performing targeted quantification of known compounds in LC-MS data using TARDIS, especially when the instrument acquired data with multiple overlapping m/z scan windows.

Use when after organism name cleaning and standardization (via 1_cleaningOriginal.R and 4_cleaningTaxonomy.R) has produced a cleaned organism table (interim/tables/2_cleaned/organism/cleaned.tsv.gz), apply this skill to map those cleaned names to Open Tree of Life (OTL) identifiers.

Use when when reproducing or auditing FIDDLE's formula prediction pipeline, or when implementing the TCN encoder in your own codebase and need to confirm that the precursor m/z (env[:, 0]) has been removed from the feature vector to avoid leakage of mass information into the model's learned.

Use when you have tandem MS data with technical replicates and need to remove features showing high variability between replicates.

Use when when you need to verify whether a specific mass spectrometry instrument platform (vendor and model), acquisition mode (e.g., targeted, ddMS2-topN, AIF, direct infusion, imaging), or file format is compatible with a lipidomics or proteomics software tool;

Use when when you have validated SMILES strings or RDKit molecule objects representing chemical structures and need to feed them into a pre-trained deep learning model (such as PS2MS, NEIMS, or DeepEI) that expects fixed-size numerical tensor inputs.

Use when after converting or downloading a pre-trained Keras model to HDF5 TensorFlow 2.3.0 format, particularly when integrating the model into a fixed-interface pipeline (e.g., NP Classifier) that expects specific named input/output layers.

Use when you have deployed a TensorFlow model via TensorFlow Serving in a containerized environment (e.

Use when you need to verify whether a GitHub Actions workflow badge (e.g., main.yml) accurately reports the CI pipeline's true pass/fail status.

Use when when you need to validate that a package's periodic integration test suite (distinct from unit tests) passes as expected, or when you must collect and communicate structured evidence of test outcomes across multiple test cases.

Use when when you need to verify that a research software package (e.g., MassQL) maintains functional correctness over time, assess the reliability of a tool before integration into a workflow, or document test coverage and failure modes for reproducibility.

Use when when you have peak-abundance .csv files with assigned molecular formulas (elemental composition: C, H, O, N, P, S) from FT-ICR MS or high-resolution MS and need to characterize the redox and structural properties of the molecular pool—e.

Use when you have downloaded or cloned a fragmentation library repository (such as LipidMatch) and need to verify that it contains the expected breadth of coverage across both molecular diversity (distinct species count) and chemical classification (lipid-type category count).

Use when when you have obtained or are considering use of the tima Docker image (adafede/tima-r) and need to confirm that the containerized environment is operational before proceeding with metabolite annotation workflows. This is a smoke test to catch environment or registry issues early.

Use when when deploying a metabolomics processing tool (such as asari) and needing to predict resource requirements or validate claimed scalability on laptop-class hardware (≤16 GB RAM, single CPU core).

Use when working with raw FT-ICR transient data (e.g., ESI_NEG_SRFA.d format) prior to noise thresholding and mass-domain calibration.

Use when you have exported lipid identifications from MS-DIAL (version 4 or 5) and need to run LipoCLEAN quality filtering on that output.

Use when you are setting up a new LipoCLEAN analysis for MS-DIAL output and need to create a configuration file tailored to your MS-DIAL version (4 or 5).

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 a ranked candidate list (e.g., BGCs sorted by IOKR or strain-correlation score) for each test spectrum, a known ground-truth BGC for each spectrum, and you want to measure retrieval performance across multiple recall depths (top-1 through top-200) and overall discrimination.

Use when after applying a configuration fix (e.g., adding an instrument type to an allowlist, updating filtering thresholds) to a dataset preprocessing pipeline, you need to confirm that the change produces the documented training/test split counts.

Use when after labeling a representative subset of peaks (typically 10–20 pooled samples with corresponding feature tables) and before neural network training, when you need to split labeled data into independent subsets for model training, hyperparameter tuning, and unbiased performance evaluation.

Use when you have RNA-seq read counts (FPKM or similar) for multiple cell lines or biological samples, a genome-scale metabolic model with GPR associations, and you need to constrain or weight metabolic reactions based on transcriptional regulation.

Use when after mass-difference network generation has matched pairwise peak mass differences to a reference biochemical transformation key with mass error ≤1 ppm, and you need to distinguish metabolic transformations driven by microbial activity from those arising from non-biological chemical.

Use when when building a neural network to map between mass spectrometry spectra and molecular properties (e.g., fingerprints, SMILES, or fragment ions) where sequential or spectral feature dependencies must be captured.

Use when after generating a Chemical Feature Tree artifact (Phylogeny[Rooted]) from q2-qemistree's make-hierarchy method, or when importing a tree from external sources, to confirm it is well-formed before proceeding to alpha/beta-diversity analysis.

Use when after applying one or more mpactr filters (filter_mispicked_ions, filter_group, filter_cv, filter_insource_ions) to an mpactr object, use this skill when you need to communicate the count and percentage breakdown of ions retained vs. rejected across filter status categories.

Use when you have a trained NeatMS neural network model (.h5 format) and need to assess its classification performance at a specific decision threshold (e.g., 0.01) to determine what fraction of true peaks are retained (TPR) versus what fraction of incorrect peaks are incorrectly accepted (FPR).

Use when when you have a precomputed similarity matrix of mass spectra (e.

Use when you have TSV or CSV files containing structure-organism pairs (with columns for structure identifier and organism identifier) and need to count unique pairs, unique structures, unique organisms, or enumerate source database identifiers to validate or report dataset scale.

Use when you have raw or processed arrival-time data from a TWIM-MS instrument and need to convert it to CCS values for comparison across experiments or biomolecular classes.

Use when you have TWIM-MS data (arrival time and m/z values) from a multi-omic sample and need to: (1) establish a CCS calibration curve from known standards, (2) assign unidentified features to biomolecular classes (e.

Use when when you need to verify that a wrapper package (e.g., rawrr) correctly bridges R and a managed .NET assembly (such as RawFileReader), specifically to confirm that internal dispatch functions can retrieve assembly location and version string before attempting actual spectral data extraction.

Use when you have measured CCS values from (LC-)IM-MS samples spiked with U¹³C labeled internal standards (e.

Use when you have trained a DNN retention time predictor and need to rank candidate metabolites for an unknown compound: the DNN outputs both point estimates and uncertainty bounds for each candidate''s RT, and you need to convert these into probabilistic scores that reflect confidence in each.

Use when you have a GC-MS results table with a Match.Factor column (representing identification confidence) and you need to understand how many distinct compounds survive at different quality cutoffs (e.g., ≥65, ≥80, ≥90).

Use when when implementing a new ComputeConverter subclass for MSMetaEnhancer that performs local chemical structure conversions using RDKit (e.g., SMILES to InChI, canonical SMILES generation).

Use when building or extending a data extraction and conversion system (such as MESSES) where tabular data is transformed via conversion directives into JSON intermediate formats and then into domain-specific schemas.

Use when when you need to confirm that a research tool or package maintains a functioning test suite, especially before adopting it for downstream analysis or before contributing modifications. Triggered by the presence of a CI workflow badge (e.

Use when when implementing or refactoring a FileInterface._open method or similar polymorphic dispatcher that conditionally instantiates different handler classes based on file extension (e.g., .gz, .db) or format metadata (e.g., indexed gzip detection).

Use when when you have an existing real mzML file from a metabolomics LC-MS/MS acquisition (e.g., beer or urine samples) and need to populate a virtual mass spectrometer with the actual chemicals that were measured, so that you can replay the acquisition with alternative fragmentation strategies (e.

Use when you have two separate LC-MS untargeted metabolomic feature tables (e.