Claude Skills by HolobiomicsLab

Use when after a Transformer model (e.g. GCMSFormer) has predicted the pure mass spectral matrix S for all components in overlapped peaks.

Use when when you have 1H-NMR metabolite measurements from Nightingale Health assayed on a new cohort and wish to compute risk scores (e.g., all-cause mortality, cardiovascular event, type 2 diabetes) using published metabolic biomarker weights from a reference study.

Use when after raw lipidomic and metabolomic data files have been generated by the Multi-ABLE method and loaded into the R environment, but before performing multivariate statistical analysis to identify differential lipids and metabolites.

Use when apply log transformation when peak intensity distributions are right-skewed with heteroscedastic variance (intensity-dependent noise), particularly in QC-based batch correction workflows where variance stabilization improves the effectiveness of subsequent Pareto scaling normalization.

Use when you have low-resolution GC-MS data (e.g., ANDI NetCDF format) with detected peaks that lack high mass accuracy, and you need to assign compound identities using a combination of retention index (RI) calibration and mass spectral library matching.

Use when when processing low-resolution GC-MS data in NetCDF format where you have already performed retention-index calibration and peak deconvolution, and you need to assign compound identities by comparing experimental mass spectra to a curated reference library such as PNNLMetV20191015.MSL.

Use when when processing GC–MS or LC–MS data as m/z vs retention time chromatograms and you need to identify biomarker or chemical marker features without conventional peak picking, particularly when false positive detection rates from peak detection algorithms are problematic.

Use when when processing raw chromatography–mass spectrometry data (GC–MS or LC–MS) as a 2D m/z vs retention time map and you need to identify and visualize marker features for analyte discrimination without relying on conventional peak picking.

Use when when processing raw FT-ICR transient data (e.g., ESI_NEG_SRFA.d) that requires assignment of molecular formulas to experimental m/z peaks. Calibration is necessary before SearchMolecularFormulas because uncalibrated mass error will cause false formula rejections or incorrect assignments.

Use when after chromatographic peak detection on preprocessed LC-MS data, when you have detected features (peaks) in multiple samples and need to establish which peaks across samples represent the same molecular species.

Use when when processing low- or high-resolution mass spectrometry data (mzML profile or centroided format) for isotopologue quantification and you need to match detected peaks to a targeted formulaTable of compounds.

Use when raw GC-MS data (netCDF or mzML format) contains overlapping chromatographic peaks with co-eluting ions that cannot be resolved by retention time alone.

Use when when you have limited real GC-MS overlapped peak data but need thousands of labeled examples to train a deep learning model for mass spectral deconvolution.

Use when you have raw LC-MS/MS spectral data in vendor formats or unvalidated .mgf files before feeding them into the specXplore importing pipeline.

Use when you have raw GC-MS data with overlapped peaks in a specific retention time region and need to resolve the individual pure mass spectra of all components present in that region.

Use when you have multiple mass spectral libraries in different formats (NIST MSP + MOL folder, MoNA MSP, RIKEN MSP, SWGDRUG MSP) and need to merge them into a single, MS-DIAL-compatible MSP file with consistent SMILES assignments, Kovats retention indices (RI), and polarity annotations across all.

Use when you have acquired EI or MS/MS spectral libraries from multiple public sources (NIST, RIKEN, MoNA, SWGDRUG, GNPS) with inconsistent metadata field layouts, missing or misplaced SMILES entries, undocumented retention indices, or mixed polarity modes, and you need to merge them into a single.

Use when when applying a pre-trained Spec2Vec Word2Vec model to new mass spectra (particularly those outside the model's training distribution), you need to assess whether peaks and neutral losses in query spectra have been seen during model training.

Use when after XCMS feature detection, grouping, and retention time correction when you have aligned features with consistent retention times and intensity patterns across samples.

Use when you have a combined EI library (from multiple sources such as NIST, RIKEN, MoNA) and access to NIST RI database files (ri.dat and USER.

Use when you have a preprocessed GC-MS dataset (from spreadOut) with standardized column names (Compound.Name, Component.RT, Base.Peak.MZ, Component.Area, Match.Factor) and a specific list of chemical compounds you want to extract and aggregate across multiple sample runs.

Use when you have two LC-MS feature tables (each containing m/z, retention time, and intensity columns) from the same or related biological samples and need to identify which features in dataset A correspond to which features in dataset B.

Use when you have mass spectral libraries from multiple sources (e.g., NIST EI, RIKEN MS2, MoNA GC-MS or LC-MS/MS, GNPS mgf) that need to be consolidated for use in MS-DIAL, or you have a single library with incomplete or malformed metadata (e.

Use when you have raw GC-MS output exported as CSV (containing columns: Component.RT, Base.Peak.MZ, Component.Area, Compound.Name, Match.Factor, File.

Use when you have received raw or vendor-converted centroid mzML files from LC-MS, GC-MS, or DI-MS platforms and need to extract MS1 spectra before building mass tracks, performing peak detection, or constructing composite feature maps.

Use when you have generated a complete feature table from mzML files (e.g., Asari 'full' feature table) and need to curate it for downstream analysis.

Use when after raw data processing and feature extraction (e.g., via XCMS, OpenMS, or enviPick) when you have detected features across multiple LC-MS or GC-MS analyses and need to identify which features represent the same chemical across samples.

Use when when you have an unknown MS/MS spectrum (with ≥10 peaks, precursor m/z, and at least 5 fragment ions) and need to identify it by comparing against a curated spectral library with annotated InChIKeys or chemical structures.

Use when when you have detected peaks in a direct injection FTICR-MS mzML file (or similar high-resolution MS format) and need to assess whether m/z measurements are accurate and consistent across the m/z range.

Use when you have a GC-MS dataset with a Match.Factor column (output from Agilent Unknowns Analysis or equivalent) and need to retain only high-confidence compound identifications.

Use when when you need to create negative control or background-only reference datasets for LC/GC-MS analysis pipelines—specifically to validate peak-picking algorithms, assess false-positive rates, or simulate instrument background and matrix effects (e.

Use when you have deconvolved GC-MS spectra (output from auto-deconvolution) and want to group similar spectra into clusters to discover spectral families, identify co-eluting compounds, or organize unknown metabolites by structural similarity.

Use when you have generated or obtained a two-dimensional mass-spectrometry intensity matrix (m/z × retention time scan points) with simulated or experimental peak shapes, noise, and background, and need to encode it as a binary .

Use when you have high-resolution LC-MS data processed through both XCMS feature detection and RAMClustR clustering, and you need to verify the reliability of molecular weight assignments before downstream annotation or statistical analysis.

Use when when you have high-resolution LC-MS or GC-MS data from environmental samples and need to simultaneously screen for both known suspect chemicals and their transformation products, rather than targeting single compounds.

Use when you have pre-processed MS/MS spectra and need to prepare them for word-embedding-based similarity methods (e.g., Spec2Vec).

Use when after feature extraction from raw LC-MS or GC-MS data (using XCMS, MS-Dial, or similar), when you have a feature intensity table with m/z and RT metadata and a reference compound database (known molecules list with m/z, RT, and annotation metadata), and you need to assign confidence-ranked.

Use when you have raw GC-MS data in netCDF or vendor-specific binary format and need to separate co-eluting compounds and extract clean mass spectra for each individual chemical component prior to molecular networking, spectral matching, or metabolite identification workflows.

Use when you have raw or preprocessed electron ionization (EI) mass spectral data that must be stored in, retrieved from, or validated against the MSP file format (used by NIST MS Search and similar spectral library tools).

Use when when you have a query electron ionization mass spectrum (as m/z and intensity pairs) and need to identify it against a spectral library stored in msp format.

Use when when you have parsed EI-MS spectrum data (m/z and intensity values) and need to feed it into a pre-trained MWFormer transformer model for direct molecular weight prediction.

Use when you have a GC-MS dataset with Match.Factor scores for each detected compound (output from Agilent Unknowns Analysis or equivalent), and you want to reduce the number of query chemicals passed to computationally intensive cheminformatics functions (categorate, mzExacto, or exactoThese).

Use when you have ESI/LC-MS test spectra requiring candidate metabolite ranking, pre-trained MLP (NEIMS) and GNN baseline models are available or can be trained, you seek quantified improvement over single-model average rank performance (baseline MLP shows ~339 average rank), and your evaluation.

Use when you have ESI/LC-MS test spectra requiring metabolite annotation and need to compare ensemble-based neural network predictions (ESP) against a baseline MLP model to quantify performance gains.

Use when after preprocessing GCxGC-MS chromatograms (smoothing, baseline correction, peak alignment) when you need to uncover latent metabolite patterns that distinguish biological groups without prior class labels. Specifically useful for exploratory separation of disease states (e.

Use when after running RAMClustR clustering on XCMS-detected LC-MS features in positive ionization mode, when you need to assign molecular weights to compound clusters and want to cross-validate the two available scoring methods (findMain and RAMClustR internal scoring) to identify cases where they.

Use when after XCMS feature detection, grouping, retention time correction, regrouping, and missing value filling on LC-MS or GC-MS data, when you have an aligned feature table with retention times and intensity profiles across multiple samples and need to collapse redundant features into.

Use when you have observed compounds (from LC-MS/MS, GC-MS, NMR, or other analytical techniques) with unknown identity and you want to assign candidate metabolite structures by comparing them to computationally predicted metabolism pathways.

Use when you have executed mzExacto() on a preprocessed GC-MS dataset and need to verify that the returned dataframe correctly matches query chemicals to their m/z peaks, retention times, and quantitative measurements (area values).

Use when after training an ensemble model (MLP, GNN, or ESP) on spectral data, use this skill to measure performance on test spectra where ground-truth metabolite identities are known. Essential for comparing model variants (e.