HolobiomicsLab
- 7.4k skills
- 0 followers
- 2 days ago last updated
- ▌ Transcription Factor Footprint Scoring · holobiomicslabUse when after bias-corrected ATAC-seq signal tracks (bigWig files) have been generated and you need to quantify transcription factor binding strength within open chromatin regions.
- ▌ Backend Performance Tradeoff Analysis · holobiomicslabUse when you are designing or optimizing an MsBackend implementation and need to decide whether to pre-populate the @spectraVars slot with all core spectra variable columns (mz, intensity, rtime, scanIndex, precursorMz, precursorIntensity, acquisitionNum, msLevel) at initialization time, or defer.
- ▌ Batch Script Execution And Automation · holobiomicslabUse when you have generated a peak table or feature list output file from an external peak-picking tool (MZmine, XCMS, MS-DIAL, or Compound Discoverer) in its native export format and need to ingest it into LipidMatch for lipid identification without manual column remapping or format conversion.
- ▌ Combination Function Parameterisation · holobiomicslabUse when you have two or more independent, standardised scoring functions that rank the same set of candidate pairs (e.
- ▌ Compositional Data Transformation Clr · holobiomicslabUse when apply CLR transformation when you have microbiome or metabolomic count data that sums to a constant across samples (relative abundance or compositional data) and intend to train supervised or unsupervised machine learning models (especially neural networks) that assume unbounded, linear.
- ▌ Converter Job Enumeration And Mapping · holobiomicslabUse when building a multi-source metadata annotation pipeline where converters are organized as dynamically discoverable subclasses in separate packages (e.g., MSMetaEnhancer.libs.converters.web and MSMetaEnhancer.libs.converters.
- ▌ Convolutional Neural Network Encoding · holobiomicslabUse when you have pairs of augmented ion images from mass spectrometry imaging data and need to generate low-dimensional representation vectors that maximize similarity between augmentations of the same image while avoiding representation collapse.
- ▌ Cross Script Consistency Verification · holobiomicslabUse when when a deep learning pipeline processes mass spectrometry spectra through multiple independent scripts (e.g., train_rescore.py, run_fiddle.py, test_caffeine.py) and a specific feature must be removed or masked to prevent the model from learning directly from a protected input (e.
- ▌ Cross Validation Benchmark Evaluation · holobiomicslabUse when you have developed a predictive model and need to compare its performance against established baselines (e.g., linear regression, Random Forest, Canonical Correlation Analysis) across multiple datasets with paired input-output features.
- ▌ Distribution Statistics Summarization · holobiomicslabUse when you have loaded a collection of molecular fingerprint vectors (such as biosynfoni fingerprints from a deposited dataset) and need to characterize their statistical and distributional properties before using them for machine learning, similarity searching, or method validation.
- ▌ Domain Context To Application Routing · holobiomicslabUse when when a user submits one or more MS/MS spectra and has declared or implied a domain context (microbial, plant, tissue, microbiome, food, or metadata aggregation), and the search must be executed against the appropriate domain-curated spectral library.
- ▌ Feature Normalization Standardization · holobiomicslabUse when after extracting and encoding molecular descriptors and structural features (atom types, bond connectivity, graph topology) from SMILES strings into a fixed-size numerical tensor, before passing the feature matrix to the PS2MS deep learning model for NPS prediction.
- ▌ Github Actions Workflow Configuration · holobiomicslabUse when when you have a Python package repository on GitHub and need to automatically verify that pull requests and commits pass unit tests and meet code quality standards before merge.
- ▌ Intensity To Concentration Conversion · holobiomicslabUse when you have raw MS intensity data paired with measurements from known concentration standards, and you need to produce absolute quantified concentration values rather than relative intensity measurements.
- ▌ Ion Mobility Mobilogram Visualization · holobiomicslabUse when when you have mass spectrometry data with ion mobility (drift time or 1/K₀) measurements as a continuous dimension and want to visualize intensity distributions across the ion mobility axis.
- ▌ Mass Spectrometry Data Column Mapping · holobiomicslabUse when you have generated a peak table or feature list from MZmine, XCMS, MS-DIAL, or Compound Discoverer in its native export format and need to ingest it into LipidMatch for lipid identification.
- ▌ Mass Spectrometry Image Preprocessing · holobiomicslabUse when when preparing raw mass spectrometry imaging (MSI) ion images for deep learning–based representation learning, especially when you need to generate augmented image pairs that preserve domain-specific artifacts (photon counting noise, missing pixels) while varying appearance.
- ▌ Matrix Directive Collation Validation · holobiomicslabUse when you have IC-FTMS measurement records in JSON format with multiple samples per metabolite assignment and you need to verify that a matrix directive with collate='assignment' correctly groups records by assignment identifier and merges sample intensity values into a single dictionary per.
- ▌ Metabolic Model Consensus Integration · holobiomicslabUse when you have draft metabolic reconstructions in SBML or standard format for multiple organisms in a microbial community (e.
- ▌ Metabolite Set Composition Validation · holobiomicslabUse when after constructing a background set for ORA in metabolomics: you have loaded an experimental detection list and a metabolomics pathway database, applied background-set construction logic, and need to confirm that the resulting background set has the correct size, composition, pathway.
- ▌ Multi Stage Dockerfile Interpretation · holobiomicslabUse when you need to validate that Docker image builds for multiple deployment variants (e.g., cli, dev, linux, windows) meet documented compressed size ranges, or when you must audit storage footprint across a multi-target build pipeline without pre-computed metrics.
- ▌ Networkx Object Generation And Export · holobiomicslabUse when after identifying statistically significant features and assigning them to structural clusters (isotopologue groups, adduct groups, cross-assay links) and correlation clusters via MamsiStructSearch.
- ▌ Neural Network Architecture Extension · holobiomicslabUse when you have a working base MPNN model (e.g., chemprop) and need to add task-specific feature processing layers (spectral, electronic, or domain features) to improve predictions on a specialized molecular property or spectrum.
- ▌ Ora Statistical Background Definition · holobiomicslabUse 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.
- ▌ Python Struct Packing For Binary Data · holobiomicslabUse when you need to serialize heterogeneous data (e.g., index integers and file offsets of different byte widths) into a compact binary format, or parse a binary file format specification that includes fixed-width field encodings.
- ▌ Quality Control Threshold Application · holobiomicslabUse when you have a feature table with intensity values across study (unknown) and blank control samples, and you need to remove features that may represent instrument artifacts, contamination, or noise rather than true biological signal.
- ▌ Spectral Library Matching And Ranking · holobiomicslabUse 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.
- ▌ Spectral Library Matching And Scoring · holobiomicslabUse when when you have MS/MS fragment spectra (in .mgf format) acquired from unknown metabolite features and need to annotate them against known compounds.
- ▌ Spectrum Plot Generation Mz Intensity · holobiomicslabUse when you have mass spectrometry spectral data loaded into a Pandas DataFrame with columns representing m/z (mass-to-charge ratio) and intensity values, and you need to visualize the spectrum to inspect peak patterns, identify high-abundance ions, or compare spectral profiles across samples or.
- ▌ Summary Statistics Table Construction · holobiomicslabUse when when you need to quantify and compare the effect of multiple filtering thresholds (e.g., Match.Factor ≥65, ≥80, ≥90) on the size of a retained compound set.
- ▌ Workflow Definition Schema Validation · holobiomicslabUse when you have located a workflow definition file (YAML or JSON) in a versioned release or commit and need to verify that it conforms to the schema specification for that release version (e.g., v1.0.0).
- ▌ Annotation Candidate Ranking Tp Score · holobiomicslabUse when after generating TP candidates from annotation algorithms (ann_comp for structure-based, ann_form for formula-based), when you need to filter and rank candidates to focus on the most credible identifications.
- ▌ Arrival Time To Drift Time Conversion · holobiomicslabUse when when processing raw TWIM-MS experimental data that contains arrival time measurements but you need drift times for CCS calibration or class-specific CCS calculations.
- ▌ Background Reference Set Construction · holobiomicslabUse when when preparing to perform pathway enrichment analysis on metabolomics data, you need a background-reference file that maps metabolic pathways to their constituent metabolites.
- ▌ Batch Effect Removal Metabolomic Data · holobiomicslabUse when after integrating feature matrices from multiple analytical experiments or batches (e.g., n=3 or more batches). Use it when batch-related systematic variations are present in the integrated dataset before proceeding to sample separation, marker identification, or annotation steps.
- ▌ Biodegradation Pathway Interpretation · holobiomicslabUse when you have a small-molecule chemical structure (as SMILES, MOL, or SDF) and need to predict how soil or aquatic microbiota will degrade or biotransform it.
- ▌ Biosynformatic Descriptor Computation · holobiomicslabUse when when you have a natural product structure (as a molecule object, SMILES string, or InChI string) and need to generate a biosynformatic descriptor for downstream machine learning tasks such as biosynthetic class prediction, structure-activity correlation, or natural product database.
- ▌ Bruker Baf Data Import And Conversion · holobiomicslabUse when you have Bruker .d/.baf format mass spectrometry imaging data and need to ingest it into MSIGen for conversion to visualizable ion images. This skill applies when your raw data originates from Bruker TIMSTOF or similar instruments and you lack direct .
- ▌ Chemical Space Similarity Computation · holobiomicslabUse when you have retrieved multiple candidate structures from a molecular structure database (e.g., PubChem, HMDB) for an unknown compound, and you have predictions of structurally related metabolites from a deep-learning mass spectral model (e.g., DeepMASS2).
- ▌ Cheminformatic Feature Representation · holobiomicslabUse when when you have natural product structures in SMILES, InChI, or SDF format and need to convert them into numerical feature vectors for downstream tasks such as biosynthetic class prediction, molecular similarity analysis, or machine learning-based natural product classification.
- ▌ Chemodiversity Descriptor Calculation · holobiomicslabUse when you have a formula-assigned FT-ICR MS dataset (CSV or tab-delimited table with molecular formulas and mass values) and need to quantify the molecular chemodiversity, elemental composition diversity, or complexity of environmental organic samples for cross-sample comparison or.
- ▌ Chromatogram Visualization Generation · holobiomicslabUse when after importing mass spectrometry data in .raw, .d, or mzXML format into R and before peak analysis or quality control steps.
- ▌ Class Conditional Calibration Mapping · holobiomicslabUse when you have TWIM-MS experimental data (arrival times and ion mobility parameters) paired with pre-assigned biomolecular class labels for an ion population, and you need to obtain class-conditioned CCS values without first performing feature-level identification.
- ▌ Classification Performance Evaluation · holobiomicslabUse when after running inference on test mass spectrometry spectra with a trained deep learning model (e.g., PS2MS) to verify that class label predictions and confidence scores match expected reference outputs.
- ▌ Compound Identifier Cross Referencing · holobiomicslabUse when when you have prepared raw mass spectrometry transition data (precursor m/z, product m/z, intensity, retention time, compound identifiers) and need to verify it conforms to EISA-EXPOSOME's required schema before database import. Specifically: you are building or validating a custom .xlsx/.
- ▌ Cross Database Nomenclature Alignment · holobiomicslabUse when when you have lipid abbreviations or names from different databases (HMDB, LIPID MAPS, LipidHome, RefMet, SwissLipids) or lipidomics software (LipidSearch, MS-DIAL, LipidBlast, etc.) that must be cross-referenced, compared, or integrated into a single annotation scheme.
- ▌ Cross Language Workflow Orchestration · holobiomicslabUse when you have multi-language code implementations (R and MATLAB scripts) for a single scientific workflow, documented example scripts for a reference sub-sample scenario, and need to verify that outputs from one language can serve as inputs to the next, or that both implementations produce.
- ▌ Dashboard Data Loading And Validation · holobiomicslabUse when after generating a dashboard_data.json file from the msFeaST Jupyter pipeline, use this skill to verify that the JSON file is correctly formatted and completely loaded into the interactive dashboard before conducting visual exploration or sharing the dashboard with collaborators.
- ▌ Differential Peak Abundance Detection · holobiomicslabUse when after raw mass spectrometry data has been converted to a peak feature table (CSV or tabular format) containing mass-to-charge ratios, retention times, and intensity values across samples, and you need to reduce noise and identify which peaks show meaningful differential patterns between.
- ▌ Domain Specific Masst API Integration · holobiomicslabUse when your research involves searching MS/MS spectra against multiple curated taxonomic or domain-specific databases (microbial, plant, tissue, microbiome, or food origin) and you need to aggregate, compare, and visualize matching results across all domains in a single interface.
- ▌ Expression Level Filtering With Edger · holobiomicslabUse when when you have a raw count matrix derived from Salmon or similar quantification tools and need to remove lowly-expressed features before normalization and batch correction.
- ▌ Expression Matrix Loading And Parsing · holobiomicslabUse when you have raw omics expression data in CSV format (rows=genes/features, columns=samples) and a separate group annotation file (sample IDs mapped to experimental groups or phenotypes), and you need to prepare them for heatmap or other omics visualizations in an R Shiny environment.
- ▌ Feature Attribution Score Calculation · holobiomicslabUse when after training a multi-layer perceptron neural network to predict metabolomic features from microbiome features using cross-validated evaluation, extract attribution scores to interpret which microbe-metabolite pairs have learned predictive relationships.
- ▌ Feature Filtering And Quality Control · holobiomicslabUse when after batch correction and concentration normalization have been applied to a merged m/z peak table and metadata file, but before statistical testing or machine learning.
- ▌ Fingerprint Feature Format Conversion · holobiomicslabUse when you have a pretrained deep learning model (e.
- ▌ Flux Variability Analysis For Scaling · holobiomicslabUse when you have a constraint-based metabolic model and need to establish per-reaction flux scaling factors derived from gene expression or other activity scores. Use FVA before integrating transcriptomics-derived constraints (e.
- ▌ Functional Level Metabolic Annotation · holobiomicslabUse when you have reconstructed the structural topology of two metabolic networks from KEGG and need to assign functional meaning (pathway annotations, metabolic functions) to each node or reaction so that you can compare the two organisms at the functional level rather than only at the topological.
- ▌ Galaxy Tool Registration Verification · holobiomicslabUse when after deploying Galaxy-M tool files and XML wrappers into a Galaxy installation's tool directories and restarting the Galaxy service, use this skill to confirm all metabolomics tools appear in the Galaxy admin interface and are ready for end-user access.
- ▌ Genbank Format Parsing And Processing · holobiomicslabUse when you have BGC sequences in GenBank format and need to extract gene-level features (coordinates, sequences, functional annotations) as input to domain annotation and tokenization pipelines. Use this skill when starting a BGC analysis from raw GenBank files rather than pre-parsed gene lists.
- ▌ Graph Embedding Coordinate Extraction · holobiomicslabUse when you have a network object (loaded as igraph or MetaNet format with edges and nodes defined) and need to assign two-dimensional spatial coordinates to nodes for downstream visualization on static platforms (ggplot2, base R graphics) or interactive viewers (Gephi, Cytoscape).
- ▌ Imaging Parameter Metadata Annotation · holobiomicslabUse when after mzML-to-imzML conversion has produced barebones imzML files with pixel alignment but no experimental metadata.
- ▌ Interactive Filter Control Design Web · holobiomicslabUse when when you have a web-based visualization of aligned mass spectrometry peaks (m/z, intensity, retention time, alignment quality metrics) and need users to interactively explore subsets of those peaks by applying constraints on intensity thresholds, alignment score cutoffs, or peak.
- ▌ Isotopologue Ratio Quality Assessment · holobiomicslabUse when evaluating the reliability of non-targeted data pre-processing (NPP) tools (XCMS, MZmine 2, MS-DIAL, etc.) on known metabolite peaks.
- ▌ Jacobian Precision Matrix Computation · holobiomicslabUse when after generating a covariance matrix from normalized metabolite abundance data (e.
- ▌ JSON Conversion Directive Application · holobiomicslabUse when you have validated intermediate JSON data conforming to the Experiment Description Specification and need to convert it to a target format (e.g., mwTab for Metabolomics Workbench deposition).
- ▌ Latent Space Dimensionality Reduction · holobiomicslabUse when you have imaging mass spectrometry (IMS) datasets where peak intensities are high-dimensional and sparse, and you need to extract compressed latent features that preserve spatial adjacency relationships and enable iterative automatic peak picking to identify marker ions.
- ▌ Lipid Abundance Differential Analysis · holobiomicslabUse when you have a preprocessed and normalized lipid abundance matrix with batch identifiers and experimental group assignments, and you need to compare lipid levels across treatment conditions (e.
- ▌ Lipid Sample Separation Visualization · holobiomicslabUse when after normalizing and log-transforming lipidomics intensity data in a LipidomicsExperiment object, use this skill when you need to assess whether your sample groups (e.
- ▌ Manifest Generation And Documentation · holobiomicslabUse when you need to audit a bioinformatics repository (such as MIBiG) to determine which sequence files are maintained locally but lack public accessions in NCBI GenBank/RefSeq. Use this to support curation workflows, data provenance tracking, or to identify candidate sequences for public release.
- ▌ Masked Modeling Pre Training Strategy · holobiomicslabUse when when you have unlabeled molecular structure data (SMILES or molecular graphs) from natural products and need to learn task-agnostic representations that capture both evolutionary (scaffold-level) and structural (side-chain) information before finetuning on downstream classification or.
- ▌ Mass Error Calculation And Assessment · holobiomicslabUse when after peak picking and before or after molecular formula assignment on FT-ICR data, especially when calibrating on a specific field strength (e.g., 12 T or 15 T).
- ▌ Mass Spectrometry Image Normalization · holobiomicslabUse when after loading a pixel array (NumPy format) and its associated metadata JSON file from MSIGen, when you need to account for pixel-to-pixel variations in total ion signal or when comparing relative abundances of multiple ions within or across samples.
- ▌ Mass Spectrometry Imaging Data Import · holobiomicslabUse when you have raw MS imaging data in imzML (continuous or processed) or Analyze 7.5 format and need to load it into R for spectral processing, normalization, peak-picking, or statistical analysis.
- ▌ Matrix Numerical Stability Validation · holobiomicslabUse when after computing a Jacobian (precision) matrix from covariance matrix inversion in MInfer, or whenever precision matrices are derived via numerical linear algebra operations on covariance matrices.
- ▌ Metabolite Abundance Matrix Filtering · holobiomicslabUse when when you have multiple batches of metabolomics data in SummarizedExperiment format with log-transformed abundance assays and need to: (1) remove poorly-quantified metabolites with >50% missing values within each batch, (2) retain only metabolites present across all batches for cross-batch.
- ▌ Metabolite Feature Annotation Mapping · holobiomicslabUse when after importing raw metabolomics data (e.g., from Metabolon, Nightingale, SomaLogic, or Olink platforms) into a Metaboprep object, but before quality control or statistical analysis.
- ▌ Metabolomic Peak Matrix Preprocessing · holobiomicslabUse when you have a raw or minimally processed metabolomic peak matrix (in standardized or tool-generated format) from multi-class samples with optional quality control (QC) samples and/or internal standards (IS), and you need to generate a single normalized peak table suitable for statistical.
- ▌ Metabolomics Data Downstream Analysis · holobiomicslabUse when you have deposited spatio-molecular matrices (e.g., MORPHnMOL.csv from SpaceM analysis) and need to reproduce or extend the data transformations, feature extractions, and figure generation reported in a metabolomics manuscript.
- ▌ Metabolomics Missing Value Imputation · holobiomicslabUse when raw metabolomics data matrices contain zero values or NA entries representing undetected metabolic features. Imputation is required before batch effect correction models (ComBat, ber) can be reliably applied, as these models assume complete feature matrices.
- ▌ Missing Value Imputation Metabolomics · holobiomicslabUse when your raw metabolomics dataset contains missing values (NAs) in metabolite columns after loading and you have already identified and removed metabolite columns with >10% missing data prevalence.
- ▌ Molecular Formula Property Derivation · holobiomicslabUse when you have filtered FT-ICR MS peak data with valid molecular formula assignments (C, H, O, N, S, P elemental counts) and need to quantify molecular properties that predict metabolite reactivity, bioavailability, and biochemical role.
- ▌ Ms Data Format Parsing And Conversion · holobiomicslabUse when you have raw breath HRMS data in mzML or mzXML format and need to extract volatile organic compound (VOC) features as a standardized CSV table indexed by m/z value, with columns for scan time or sample identifiers and corresponding intensity measurements.
- ▌ Msi Data Processing Speed Measurement · holobiomicslabUse when when you need to validate that MSI software (e.g., LipidQMap) achieves documented processing speeds on your target hardware, or when you need to establish a performance baseline before deploying the software for high-throughput imaging studies.
- ▌ Msimagingexperiment Object Validation · holobiomicslabUse when after reading an imzML file (continuous or processed format) using readMSIData() and before proceeding to preprocessing or statistical analysis steps. Use this skill whenever you need to confirm that a parsed imaging dataset meets expected structural requirements—e.
- ▌ Msimagingexperiment Subset Operations · holobiomicslabUse when after peak alignment with peakAlign(), when you have an MSImagingExperiment with many detected peaks but want to retain only those present in a sufficient fraction of pixels.
- ▌ Multi Criteria Performance Evaluation · holobiomicslabUse when you have preprocessed multi-class or time-course metabolomic peak tables (with or without quality control samples and/or internal standards) and need to compare multiple preprocessing workflows to identify which performs for biomarker discovery or differential profiling.
- ▌ Multi Fold Split Validation Reporting · holobiomicslabUse when when evaluating a spectral embedding or compound identification model on a dataset where the training/test split has already been finalized (e.
- ▌ Multi Format Identifier Harmonization · holobiomicslabUse when you have metabolomics datasets from multiple studies that use different chemical identifier formats (names, InChI, InChIKey, SMILES), and you need to merge them for meta-analysis or cross-study comparison.
- ▌ Natural Isotope Abundance Calculation · holobiomicslabUse when when processing mass spectrometry imaging (MSI) data in positive ion mode where both [M+H]+ and [M+Na]+ adducts are present for the same lipid species, and you observe intensity overlap in [M+H]+ ion images caused by the isotopic fine structure of [M+Na]+ adducts.
- ▌ Neural Network Encoder Implementation · holobiomicslabUse when when you need to benchmark multiple encoder types (e.g., FFN vs. GNN) on the same predictive task and require evidence that performance differences reflect genuine architectural trade-offs rather than suboptimal tuning.
- ▌ Neural Network Model Layer Validation · holobiomicslabUse when after converting Keras models to HDF5 TensorFlow 2.0 format and before deploying them to a TensorFlow Serving endpoint.
- ▌ Neutral Mass Inference Via Regression · holobiomicslabUse when use this skill after khipu has assigned observed ions to grid positions (isotope and adduct combinations). Apply it when you have a connected subnetwork of feature ions that have been matched to known isotope and adduct patterns and need to estimate the neutral mass of the parent compound.
- ▌ Nonparametric Reproducibility Ranking · holobiomicslabUse when you have high-dimensional replicate experiment data (e.
- ▌ Over Representation Analysis Workflow · holobiomicslabUse when when you need to reproduce published ORA simulation results, validate pathway enrichment findings from a metabolomics study, or examine pitfalls and practices in Over-representation Analysis methodology.
- ▌ Pandas Dataframe Column Specification · holobiomicslabUse when when you have mass spectrometry data in a Pandas DataFrame with column names that do not match pyOpenMS-viz's default expectations (e.g., 'm/z' vs 'mz' or 'retention_time' vs 'rt'), or when your data uses domain-specific column labels (e.g., 'mass_to_charge', 'scan_time', 'peak_intensity').
- ▌ Parameter Sharing In Siamese Networks · holobiomicslabUse when when processing paired augmented versions of the same input (e.g., two augmented ion images in COL or ISO mode) and you need to learn meaningful low-dimensional representations via contrastive loss.
- ▌ Peak Background Binary Classification · holobiomicslabUse when you have LC-HRMS profile-mode chromatograms with extracted local maxima exported as standardized 2D rt×mz areas, and you need to disambiguate true chromatographic peaks from background signals (including wall artifacts and noise) at scale.
- ▌ Peak Detection Parameter Optimization · holobiomicslabUse when when loading and processing raw or recalibrated FT-ICR mass spectrum data (Bruker .d format) on a defined field-strength instrument (e.g., 12 T or 15 T), before executing molecular formula search.
- ▌ Phase I Phase Ii Transformation Rules · holobiomicslabUse when when you have a parent drug's raw chemical formula and need to predict its likely metabolite formulas for subsequent mass spectrometry matching, especially in high-resolution MS workflows where exact mass matching requires known candidate formulas.
- ▌ Phenotype Pathway Association Ranking · holobiomicslabUse when after computing a pathway dysregulation score matrix (PDSmatrix) from metabolite-pathway associations, apply this skill when you need to reduce the pathway feature space to those most predictive or explanatory of a phenotype label.