HolobiomicsLab
- 7.4k skills
- 0 followers
- 14 hours ago last updated
- ▌ Neutral Loss Formula Computation 2 · holobiomicslabUse when you have tandem mass spectra with precursor m/z and observed fragment peak m/z values (as mz/intensity pairs), and you need to construct interpretable feature vectors where each axis corresponds to a real chemical entity (peak or neutral loss) rather than a latent dimension.
- ▌ Nmr Spectrum Object Construction 2 · holobiomicslabUse when you have Bruker NMR spectral files (raw instrumental output) and need to prepare them for automated metabolite identification and quantification.
- ▌ Peak Deconvolution Preprocessing 2 · holobiomicslabUse when you have raw IM-MS data in Agilent MassHunter (.d) or UIMF format from drift tube (DT) or SLIM instruments, and you intend to perform HRdm demultiplexing or peak deconvolution to resolve co-eluting or structurally similar ions.
- ▌ Peak Quality Threshold Filtering 2 · holobiomicslabUse when after composite-map peak detection (scipy.signal.find_peaks) has identified candidate peaks on aligned mass tracks, but before compiling the final feature table.
- ▌ Qc Sample Reliability Evaluation 2 · holobiomicslabUse when after drift correction has been applied to your LC-MS peak table and you need to identify low-quality metabolic features that exhibit high internal spread (RSD, RSD*) or excessive QC-versus-biological variation (D-ratio) before imputation and batch correction.
- ▌ Qc Workflow Component Inspection 2 · holobiomicslabUse when you have acquired a versioned QC workflow definition file (YAML or JSON) from a metabolomics QC system release (e.g., v1.0.
- ▌ Replicate Consistency Assessment 2 · holobiomicslabUse when after NMR or MS data acquisition and preprocessing (phasing, baseline correction) when you have a SummarizedExperiment object containing assay intensity matrix with QC sample columns designated.
- ▌ Spearman Correlation Computation 2 · holobiomicslabUse when : (1) you have metabolomic data (NMR or MS-derived) and a continuous phenotype variable; (2) you need to quantify associations while controlling for known confounders (age, gender, disease status);
- ▌ Spectral Data Quality Assessment 2 · holobiomicslabUse 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.
- ▌ Spectral Peak Embedding Encoding 2 · holobiomicslabUse when when you have variable-length MS/MS peak lists (m/z and intensity arrays) that must be fed into a transformer-based model for spectra analysis, and you need deterministic, normalized embeddings that preserve peak frequency information across multiple scales.
- ▌ Summarized Experiment Subsetting 2 · holobiomicslabUse when when you have a SummarizedExperiment containing metabolomic abundances and a corresponding vector of quality metrics (e.g., coefficient of variation computed across QC samples), and you need to filter to retain only features meeting a reproducibility threshold (e.g., CV ≤ 0.
- ▌ Reproducibility Gap Detection 2 · holobiomicslabUse when when evaluating a computational method described in a peer-reviewed article, particularly when the work references a public repository or provides supplementary code.
- ▌ M Z Intensity Calibration 2 · holobiomicslabUse when you have raw or processed MS spectrum data (m/z and intensity pairs) from DI-MS, ASAP-MS, or other high-throughput mass spectrometry instruments that requires automated peak detection.
- ▌ Artifact Removal In Ims Data 2 · holobiomicslabUse when processing raw IM-MS data (UIMF or Agilent MassHunter .d format) that exhibits jagged peaks in low-abundance ions, isolated high-intensity noise spikes, or saturated detector signals that distort elution and mobility profiles.
- ▌ Protonation State Assignment 2 · holobiomicslabUse when you have SMILES-encoded molecular structures and need to model their behavior under electrospray ionization (ESI) or other ionization methods in mass spectrometry.
- ▌ Smiles Structure Preparation 2 · holobiomicslabUse when you have a SMILES input file of small organic molecules and need to predict their collision cross sections or other molecular properties via quantum mechanics.
- ▌ Batch Spectrum Quality Control 2 · holobiomicslabUse when when ingesting spectra from multiple open mass spectrometry libraries (OMSLs) in .mgf, .msp, .json, or .csv format and you observe mixed experimental protocols, inconsistent adduct annotations, or partial metadata.
- ▌ Vocabulary Coverage Validation 2 · holobiomicslabUse when after merging separate vocabularies for distinct data modalities (e.g., spectral tokens for m/z values and intensities, structural tokens for SMILES or graphs) and before deploying the unified vocabulary in a language model.
- ▌ Mass Error Tolerance Filtering 2 · holobiomicslabUse when when annotating full-scan MS or MS imaging data against a metabolite database (e.g., LipidMaps, HMDB) and you need to control the stringency of m/z matching. Use this filter to balance annotation sensitivity against specificity: tighter ppm tolerances (e.
- ▌ Mass Spectrometry Data Extraction 2 · holobiomicslabUse when after peak picking, sample alignment, and isotopologue/adduct grouping steps have been completed in an untargeted LC-MS workflow.
- ▌ Mass Spectrometry Data Processing 2 · holobiomicslabUse when you have raw mass spectrometry outputs (peak areas/heights across samples and fragmentation spectra) that need to be formatted and validated before running the tima taxonomically informed annotation workflow.
- ▌ Mass To Charge Tolerance Matching 2 · holobiomicslabUse when you have an untargeted metabolomics feature table (with m/z and retention time columns) and a metabolic network database with compound chemical formulas, and you want to connect observed features to known metabolic reactions and pathways using mass matching rather than metabolite.
- ▌ Spectral Metadata Standardization 2 · holobiomicslabUse when when importing mass spectrometry spectral data from public repositories or multi-source MGF/MSP files where metadata fields are inconsistent, missing, or non-standard;
- ▌ Bayesian Meta Learning Projection 2 · holobiomicslabUse when you need to transfer retention time predictions from one chromatographic method to another, but have access to only a small number (≥10) of molecules with known retention times in both methods.
- ▌ Composite Mass Track Construction 2 · holobiomicslabUse when when processing a multi-sample LC-MS metabolomics project after mass-track extraction and retention-time calibration have been applied to all individual samples, and you need to detect peaks across the entire cohort.
- ▌ Compound Structure Representation 2 · holobiomicslabUse when when you have experimental MS/MS data (peak lists, precursor m/z, charge state, adduct type) paired with a chemical structure (SMILES or structural identifier), and need to create a unified Compound object for spectral alignment, modification-site prediction, or comparative fragmentation.
- ▌ Cosine Similarity Ranking Metrics 2 · holobiomicslabUse when when you have pre-computed spectral embeddings (vectors) for both query spectra and a reference library, and you need to measure retrieval performance by ranking reference compounds by similarity to each query.
- ▌ Count Verification And Validation 2 · holobiomicslabUse when you have grouped unique 2D chemical structures by organism prevalence and need to confirm that the counts in each frequency bin (singleton, low-diversity, medium-diversity, high-diversity) match published or curated reference values.
- ▌ Custom Metabolite Set Integration 2 · holobiomicslabUse when you have a user-supplied metabolite set file (CSV or JSON) defining custom groupings of metabolites (e.
- ▌ Dependency Requirement Validation 2 · holobiomicslabUse when before launching the DaDIA pipeline or any multi-package R workflow that has strict version constraints. Use this skill when you have access to an R environment and need to confirm that R ≥4.0, XCMS ≥3.11.4, and metaMS ≥1.25.
- ▌ Differential Metabolite Detection 2 · holobiomicslabUse when you have normalized and aligned lipidomic and metabolomic spectral features from the Multi-ABLE method across multiple biological samples grouped by phenotype (e.
- ▌ Endianness Detection And Handling 2 · holobiomicslabUse when when reading a binary file format with a magic integer or fixed checksum field at a known offset, and endianness is not explicitly declared in file metadata or header comments.
- ▌ Fuzzy Analog Search Fragmentation 2 · holobiomicslabUse when when you have experimental MS/MS spectra and want to discover structurally similar compounds beyond exact spectral library matches—particularly useful for identifying chemical analogs, homologs, or isomers that share fragmentation logic but differ in molecular structure.
- ▌ Github Actions Workflow Execution 2 · holobiomicslabUse when you have a GitHub repository containing scientific records (e.
- ▌ Image Processing For Metabolomics 2 · holobiomicslabUse when you have GC–MS or LC–MS data represented as a two-dimensional map (m/z vs retention time) and need to identify analyte signals and marker features while minimizing false peak detections.
- ▌ Ion Image Quantification Workflow 2 · holobiomicslabUse 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.
- ▌ Isotope Labeling Data Integration 2 · holobiomicslabUse when you have LC-MS peak tables from parallel unlabeled and labeled (isotope-traced) sample cohorts, sample metadata defining groups and conditions, and you seek to identify metabolic intermediates that accumulate differentially in a perturbed system (e.
- ▌ Lcms Spectral Peak Classification 2 · holobiomicslabUse when you have raw LC-MS spectral peak data (in the format provided by DOI 10.25345/C5FD2F) and need to build a classifier that can distinguish valid peaks from false positives or background noise without manual feature engineering.
- ▌ M Z Metabolite Annotation Mapping 2 · holobiomicslabUse when when you have a spatial metabolomics or LC-MS dataset with detected m/z features (as a feature matrix or SpaMTP Seurat object) and need to assign metabolite identities. Specifically: you have observed m/z values, you know the ionization polarity and expected adduct form (e.
- ▌ Mass Spectrometry Database Search 2 · holobiomicslabUse when you have an unknown mass spectrum (or a representative metabolite spectrum from public data) and need to identify it by comparing it against a large reference library—particularly when the database contains billions of spectra and earlier tools like MASST are too slow or resource-intensive.
- ▌ Mass Spectrometry File Validation 2 · holobiomicslabUse when after MSConvert has converted vendor-specific raw mass spectrometry data (ThermoFisher, Agilent, or equivalent formats) on a Linux system and before initiating analysis in MSThunder.
- ▌ Mass Spectrometry Query Semantics 2 · holobiomicslabUse when when you need to express complex mass spectrometry search patterns (e.g., isotope envelope detection, neutral loss patterns, intensity relationships across m/z ranges) in a human-readable format that can be executed against spectral data files (mzML, mzXML) or spectral repositories.
- ▌ Mass To Charge Tolerance Matching 3 · holobiomicslabUse when you have statistically significant LC-MS features and need to group them into structural clusters. Specifically, use it after selecting features by p-value threshold (e.g., p < 0.
- ▌ Mass Track Extraction And Binning 2 · holobiomicslabUse when when you have centroid mzML files from LC-MS metabolomics acquisition and need to construct sample-level mass tracks before cross-sample alignment. Specifically: you are starting fresh with vendor-converted or pre-processed mzML input;
- ▌ Metabolic Parameter Visualization 2 · holobiomicslabUse when when you have paired NMR metabolite measurements and corresponding processing metadata (pre-centrifugation delay, post-centrifugation delay, sample type, cohort) for a blood sample cohort and need to determine which metabolites remain stable across the expected or observed delay range, or.
- ▌ Metabolomic Heatmap Visualization 2 · holobiomicslabUse when after completing feature annotation and reaction assignment in an untargeted metabolomics workflow, specifically when you have a feature-by-sample intensity matrix aligned with metabolite identities and want to communicate cluster structure, reaction pathway groupings, and feature.
- ▌ Ms Ms Spectral Data Preprocessing 2 · holobiomicslabUse when when you have MS/MS spectral data (raw or intermediate format) that must be fed into the Mass2SMILES Docker container or similar deep learning models for MS/MS-to-structure inference.
- ▌ Msconvert Parameter Configuration 2 · holobiomicslabUse when you have acquired raw mass spectrometry data from ThermoFisher, Agilent, or compatible vendors in their native formats (.raw, .d, or equivalent) and need to prepare it for nontargeted analysis using MSThunder on a Linux system.
- ▌ Msi Feature Metadata Preservation 2 · holobiomicslabUse when you are converting a processed Cardinal MSImagingExperiment object (containing normalized peaks, optional spatial shrunken centroids segmentation, and feature m/z annotations) into a Seurat object for downstream pathway analysis, differential metabolite expression, or integrative.
- ▌ Pathway Rank Stability Assessment 2 · holobiomicslabUse 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.
- ▌ Peak Picking Algorithm Comparison 2 · holobiomicslabUse when you have claims in a paper or tool documentation that one peak picking method outperforms others (e.g., 'IDSL.IPA outperforms MZmine 2 and xcms'), but the specific comparison metrics, numerical results, and source tables are not provided in the abstract or introduction.
- ▌ Peak Table Filtering Metabolomics 2 · holobiomicslabUse when after generating a peak table from XCMS peakTable() output in an untargeted LC-MS metabolomics workflow, if your experimental design includes quality control (QC) samples (SampleType='LQC') and you want to exclude noisy or unstable EICs before building a peak quality classifier.
- ▌ Ppm Tolerance Scoring And Ranking 2 · holobiomicslabUse when you have extracted a list of candidate molecular formulae for a given m/z value and need to rank them by plausibility.
- ▌ Precursor List Formatting For Dda 2 · holobiomicslabUse when you have generated a lipid spectral library (lipid identities, adducts, m/z values, fragmentation patterns) and your downstream analysis requires DDA acquisition on an Orbitrap instrument using Excalibur software.
- ▌ Public Database Query Integration 2 · holobiomicslabUse when you have an experimental MS/MS spectrum (m/z and intensity pairs with known precursor m/z) and need to identify the compound by searching against public repositories or a local reference library.
- ▌ Quality Control Metric Evaluation 2 · holobiomicslabUse when you have measured IM-MS lipidomics data spiked with U13C labeled internal standards and need to assess whether CCS bias remains within acceptable limits for each lipid class-adduct combination.
- ▌ R Script Validation And Execution 2 · holobiomicslabUse when when you have obtained an R-based bioinformatic program (such as DNMS2Purifier.
- ▌ R Statistical Model Serialization 2 · holobiomicslabUse when after training a customized R statistical or machine learning model on annotated training data, you need to persist the trained model object for reuse in downstream analysis workflows without retraining.
- ▌ Retention Time Prediction Scoring 2 · holobiomicslabUse when you have a list of candidate metabolites for an unknown compound (from mass-to-structure search or library matching), experimental retention time(s) from one or more chromatographic methods, and access to a trained DNN RT predictor and meta-learned RT projection model.
- ▌ Signal To Noise Ratio Computation 2 · holobiomicslabUse when after peak detection in nontargeted LC-MS workflows when you have a feature table with detected peaks and need to assign quality scores or filter low-confidence features.
- ▌ Small Molecule Structure Matching 2 · holobiomicslabUse when you have observed compounds (from LC-MS, GC-MS, or spectroscopy) and a set of predicted metabolite structures from BioTransformer, and need to assign identities to the observed compounds by matching their experimental features (mass-to-charge ratio, retention time, spectral signature) to.
- ▌ Spectral Database Query Execution 2 · holobiomicslabUse when when you have an unknown mass spectrum (query spectrum) and need to search it against a reference database of billions of spectra to find matching or structurally related compounds.
- ▌ U13c Labeled Lipid Identification 2 · holobiomicslabUse when you have measured CCS values from (LC-)IM-MS samples spiked with U¹³C labeled internal standards (e.
- ▌ Web Service Deployment Validation 2 · holobiomicslabUse when when you need to confirm that a publicly hosted academic web service (such as molDiscovery) is live and responding at a documented endpoint URL, or when troubleshooting access issues reported by end users.
- ▌ Ccs Reference Data Extraction 2 · holobiomicslabUse when you have obtained or need to prepare a DTCCS_N2 reference library for U13C-labeled lipids (typically provided as part of a lipidomics tool distribution) and need to extract, validate, and normalize its contents into a machine-readable table format before using it for CCS bias calculation.
- ▌ Conformer Ensemble Processing 2 · holobiomicslabUse when you have a set of conformers that have already been filtered by ASE-ANI neural network potentials and need to extract quantum-mechanical electronic properties (polarizability tensor, dipole moment) required for collision cross section calculation.
- ▌ Per Lipid Class Normalization 2 · holobiomicslabUse when you have IM-MS lipidomic data from samples spiked with U13C-labeled internal standards (e.g., fully labeled yeast extract), measured CCS values stratified by lipid class and adduct type, and access to the DTCCS_N2 reference library for U13C lipids.
- ▌ Saturated Peak Reconstruction 2 · holobiomicslabUse when processing IM-MS data files (Agilent .d or UIMF format) that contain high-abundance ions suspected of signal saturation, particularly in untargeted or discovery proteomics/metabolomics workflows where dynamic range compression would obscure quantitative relationships.
- ▌ Smiles Canonicalization Rdkit 3 · holobiomicslabUse when when processing raw SMILES strings from external databases or user input that may contain non-canonical tautomeric forms, variable stereochemical notation, or redundant representations of the same chemical structure.
- ▌ Chromatographic Peak Processing 2 · holobiomicslabUse when after peak detection when you have a table of detected peaks with m/z values and retention times from LC/HRMS data, and you observe systematic m/z drift across a batch or population-scale study (n > 500 samples).
- ▌ Lc Hrms 2d Area Standardization 2 · holobiomicslabUse when after detecting local-maxima in LC-HRMS profile mode datasets and before training or inference with a CNN model for peak classification.
- ▌ Mass Spectrometry Preprocessing 2 · holobiomicslabUse when when you have raw mass-spectrometry data (precursor m/z, ionization mode, and fragment m/z–intensity pairs) that must be fed into a CNN model for metabolite annotation via compound fingerprint prediction.
- ▌ Search Performance Benchmarking 2 · holobiomicslabUse when when you have two or more implementations of a spectral search tool (e.g., MASST vs. MASST+) and need to quantify whether claimed performance improvements (e.g., '100-fold speedup') are reproducible.
- ▌ Unified Vocabulary Construction 2 · holobiomicslabUse when when you have parallel mass spectra and molecular structure data (e.g., CANOPUS or MassSpecGym datasets) and aim to train a single encoder-decoder model (e.g., BART) that must handle both modalities as input and output tokens.
- ▌ Binary Spectral Data Extraction 2 · holobiomicslabUse when you have parsed imzML XML metadata and loaded the corresponding .ibd binary intensity file, and need to extract specific ion images at one or more target m/z values.
- ▌ Bioinformatic Object Conversion 2 · holobiomicslabUse when when you have processed Cardinal MSI data (normalized peak intensities, optional SSC segmentation results) and need to transition to Seurat-based workflows for differential expression, pathway analysis, or integration with spatial transcriptomics data.
- ▌ Matlab Workspace Initialization 2 · holobiomicslabUse when when you have mass spectrometry imaging root datasets paired with accompanying .mat workspace files (as in the B73 and Oaxacan Green genotypes from Sama et al.
- ▌ Microbe Compound Identification 2 · holobiomicslabUse when you have collected MS/MS spectra from a microbial sample (pure culture, environmental isolate, or mixed community) and need to assign chemical identities to observed m/z features while simultaneously resolving which microbial taxa are likely producers of each metabolite.
- ▌ Mass Spectrometry Library Matching 2 · holobiomicslabUse 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.
- ▌ Attention Mechanism Implementation 2 · holobiomicslabUse when you have imaging mass spectrometry (IMS) datasets with peak intensity features organized as spatial graphs (nodes = pixels/voxels, edges = spatial adjacency), and you need to discover latent peak patterns for automatic peak picking or marker ion identification without manual feature.
- ▌ Chemical Transformation Annotation 2 · holobiomicslabUse when after computing all pairwise mass differences from MS imaging peaks and binning them into a histogram, use this skill when you need to prioritize which mass differences are most likely to represent real chemical adducts (rather than noise or measurement artifacts) by ranking them by.
- ▌ Compound Identification Validation 2 · holobiomicslabUse when when you have an unknown MS/MS spectrum (in .
- ▌ Conditional Dispatch Routing Logic 2 · holobiomicslabUse when a mass spectrometry analysis pipeline must accept data from multiple sources with different identifier schemes (GNPS Task ID, Universal Spectrum Identifiers, or Feature-Based Molecular Networking task IDs), and you need to transparently route each to the correct loader without requiring.
- ▌ Controlled Vocabulary Term Mapping 2 · holobiomicslabUse when you have collected or inherited sample-information metadata from multiple sources (e.
- ▌ Dataset Size Threshold Enforcement 2 · holobiomicslabUse when when you have partitioned public MS/MS files from MassIVE using the ReDU File Selector into one or more filtered groups (G1–G6) and need to verify that each group's file count complies with computational constraints before submitting to GNPS molecular networking (3000 file limit) or.
- ▌ Deep Learning Model Initialization 2 · holobiomicslabUse when you have downloaded the LC-MS spectral peak dataset (DOI 10.25345/C5FD2F) and need to train a DNN model from scratch rather than using a pre-trained checkpoint.
- ▌ Encoder Output Dimension Reduction 2 · holobiomicslabUse when you have 512-dimensional representation vectors output from paired ResNet18 encoders processing augmented ion images, and you need to: (1) introduce an intermediate projection space to enable contrastive loss optimization without trivial/collapsed solutions, (2) further compress learned.
- ▌ Feature Quality Assessment Metrics 2 · holobiomicslabUse when after nontargeted peak detection and segmentation has generated a feature table from raw LC-MS data (mzML or vendor format), apply quality assessment when you need to rank or filter features by confidence before annotation, adduct grouping, or MS/MS matching.
- ▌ Feature Table Row Count Validation 2 · holobiomicslabUse when after peak quality filtering has been applied to a composite map peak detection output using SNR (>2), goodness-of-fit (peakshape > 0.5), minimum peak height (default 1e5), and prominence (≥20% of peak_height) thresholds.
- ▌ File Grouping By Sample Attributes 2 · holobiomicslabUse when you have a validated ReDU sample-information metadata table (gnps_metadata.tsv) loaded from a MassIVE accession, and you need to partition public MS/MS files into multiple analysis cohorts by one or more sample attributes.
- ▌ Formula Network Extraction From Mz 2 · holobiomicslabUse when you have an observed m/z value from spatially-resolved mass spectrometry imaging and need to assign one or more plausible molecular formulae with confidence metrics.
- ▌ Github Release Metadata Comparison 2 · holobiomicslabUse when you have reproduced a release artifact locally (e.g., via Semantic Release or a build tool) and need to verify it matches the official GitHub release record.
- ▌ Gnps Molecular Network Integration 2 · holobiomicslabUse when you have computed frequent fragmentation patterns from a collection of MS/MS spectra using mineMS2, and you want to focus pattern interpretation on subsets of spectra that form meaningful network components (connected groups, cliques, or high-similarity pairs) in a GNPS molecular network.
- ▌ Gradient Space Optimization Search 2 · holobiomicslabUse when after fitting a Gaussian Process regression model to prior LC-MS gradient runs (retention times, separation efficiency scores, or compound identification counts).
- ▌ Hyperparameter Sweep Configuration 2 · holobiomicslabUse when when implementing multiple competing model architectures (e.
- ▌ Knowledge Graph Integration Design 2 · holobiomicslabUse when designing a metabolite annotation workflow that must simultaneously leverage established biochemical knowledge (pathway databases, reaction networks) and experimental evidence (mass spectrometry feature similarity, co-occurrence patterns).
- ▌ Lc Hrms Data Processing Evaluation 2 · holobiomicslabUse when you have processed the same set of untargeted LC/HRMS files (mzXML, mzML, or netCDF format) with two or more peak-picking tools and need to validate which tool produces higher-quality peaklists for organic small molecules.
- ▌ Mass Grid Construction And Mapping 2 · holobiomicslabUse when after mass track extraction from individual LC-MS samples, when you need to align mass tracks across a cohort to produce a unified feature matrix. Specifically: when study size is ≤10 samples, use pairwise anchor-prioritized alignment;
- ▌ Metabolite Feature Quality Control 2 · holobiomicslabUse when you have a metabolomic SummarizedExperiment object with replicate QC (quality control) samples and need to remove non-reproducible metabolic features before phenotype association modeling. Use it specifically when your workflow requires FDA-compliant reproducibility thresholds (CV < 0.
- ▌ Metabolite Set Decomposition Plage 2 · holobiomicslabUse when you have peak intensity data from metabolomics experiments with annotated metabolites assigned to known groupings (KEGG pathways, Reactome, GNPS Molecular Families, or MS2LDA Mass2Motifs) and need to identify which metabolite sets change significantly across experimental comparisons while.