Claude Skills by HolobiomicsLab

<!-- Use this when YOU are the author of the tool (or method/paper) a skill grounds on, and you want to contribute that skill AND be credited as a co-author. Keep it short: tell us the tool/paper, what the skill does, and who you are. A maintainer runs the `claim-skill` command, which normalizes/matches/grounds the skill (the propose-skill flow), writes you into the staged skill's `contributors` as `role: author`, and stages a reviewable PR with `Co-authored-by:` you. A maintainer makes the f...

<!-- Keep this short. A meta-skill (super-skill) does NOT re-implement a procedure — it SEQUENCES existing sub-skills into a canonical end-to-end pipeline (e.g. "molecular networking": feature detection/alignment → spectral similarity/networking → annotation/propagation → visualization). Tell us WHAT pipeline it captures, WHICH sub-skills/tools it should orchestrate, and WHY it's worth shipping as a super-skill. The `synthesize-meta-skill` command does the synthesis, grounding, and normalizat...

<!-- Keep this short. Tell us WHAT the skill does, WHY it's worth adding, and give a LINK to the skill (or paste its SKILL.md). The `propose-skill` command does the matching, grounding, and normalization, and stages a reviewable PR. A maintainer makes the final merge decision — you cannot self-merge. See governance/COMMUNITY_SKILLS.md for how community skills are curated. --> **What does the skill do?** <!-- one line, e.g. "Use when aligning RT across batches with method X" --> **Why it matte...

> 🧩 **What this is:** a super-skill is a *composite workflow* — an ordered DAG of > **existing leaf skills** for an end-to-end task (e.g. `mzML → annotation`). It is > generated from a short **skeleton** (stages + goals + preferred leaf skills) by > the ASB factory, not hand-written. It reuses leaves; it does not add new ones. > To add a *new tool/paper*, file a [Propose a paper](./propose-paper.md) instead.

Use when you have raw .idat files or beta-valued matrices from HumanMethylation450 (450k) arrays and need to remove low-quality probes, correct for technical artifacts (batch effects, type-2 probe bias), and normalize the data before performing differential methylation analysis or DMR detection.

Use when when working with large single-cell ATAC-seq or multi-omics datasets where in-memory storage is infeasible (>1M cells), and you need to iteratively add or modify count matrices (tile-based, peak-based, or gene-based) while preserving fragment-level data for reproducibility and re-analysis.

Use when you have a processed ArchR project object (containing peak calls, cell barcodes, and quality control metadata) and need to: (1) reduce dimensionality of the peak matrix via iterative LSI; (2) integrate multimodal data by combining ATAC and RNA dimensions;

Use when when you have aligned ATAC-seq BAM files and need to quantify Tn5 transposase insertion patterns around specific genomic coordinates (motif sites, peaks, regulatory regions) to detect transcription factor occupancy footprints or compare chromatin accessibility between bound and unbound.

Use when when you need to assess whether a given ATAC-seq clustering method (or variant) is competitive on your data or when evaluating which published method to adopt.

Use when when you have processed scATAC-seq data (peak calling complete, cell-barcode matrix generated) and need to register it into ArchR for downstream multiome analysis alongside scRNA-seq gene expression data.

Use when you have completed Tn5 insertion bias correction on ATAC-seq reads and now need to quantify footprint signal strength (signal depletion around TF-bound sites) across accessible chromatin regions before classifying individual TF binding sites.

Use when after differential peak analysis (tl.diff_test) has identified peaks that differ in accessibility across cell types or conditions.

Use when you have aligned ATAC-seq BAM files and need to detect transcription factor binding sites via footprint analysis. The skill is essential because raw Tn5 insertion signal contains systematic bias toward certain DNA sequences;

Use when you have raw ATAC-seq BAM files and need to perform footprinting analysis to detect transcription factor binding through Tn5 insertion patterns.

Use when after computing expected accessibility from filtered peak and sample counts, and before computing final deviation scores. Use this skill when working with sparse ATAC-seq or DNase-seq data where GC bias and accessibility depth are known confounders of motif-associated variability.

Use when before invoking pp.make_fragment_file on a BAM file from alignment or external sources, especially when the BAM's sort order is unknown or when integrating BAM files from multiple sequencing platforms (10X, standard genomics pipelines, or custom aligners) into a unified SnapATAC2 analysis.

Use when after running TOBIAS ATACorrect to generate bias-corrected signal tracks from aligned ATAC-seq reads. Use this skill when you have corrected cutsite signal (as .

Use when when you have coordinate-sorted BAM files from single-cell ATAC-seq experiments (e.g., 10X Genomics scATAC-seq) and need to generate a compressed fragment file for efficient downstream analysis. Use this step before matrix generation, clustering, or peak calling in SnapATAC2.

Use when you have a coordinate-sorted BAM file from a single-cell ATAC-seq experiment (especially 10X Genomics platforms) and need to extract per-fragment information including cell barcodes, fragment coordinates, and quality metrics for downstream analysis in SnapATAC2 or compatible tools.

Use when apply this filter after loading raw .idat files or beta-valued matrices from HumanMethylation450 or EPIC methylation arrays when you need to remove probes with insufficient bead counts that may introduce measurement noise or bias into downstream differential methylation or enrichment.

Use when you have extracted quantitative genomic features (e.g., insulation scores, boundary annotations) as a pandas DataFrame with bin coordinates and boolean or numeric columns, and need to export them as BED format for visualization in genome browsers (e.

Use when after computing per-bin coverage depth using cooltools.coverage() on a loaded cooler object, when you need to (1) share the coverage track with non-Python tools, (2) visualize it in a genome browser, or (3) integrate it with downstream analyses that expect bedGraph or tabular input.

Use when you have aligned ChIP-Seq reads (in BED or BEDPE format) and need to convert them into quantitative genome-wide signal tracks (coverage, p-value, or q-value scores) for downstream statistical comparison or peak detection.

Use when you have generated or received bedGraph files from paired-end sequencing (via bedtools genomecov or similar) and need to verify they conform to UCSC bedGraph format before passing them to peak-calling tools like SEACR.

Use when your ChIP-Seq input is paired-end sequencing data stored in BEDPE format (e.g., CTCF_PE_ChIP_chr22_50k.bedpe.gz), and you need to estimate fragment length and call peaks while respecting the paired nature of the reads rather than treating them as independent single-end alignments.

Use when you are reproducing a comparative benchmarking claim (e.

Use when after computeDeviations has generated a SummarizedExperiment object with z-score assays reflecting bias-corrected deviations of observed vs. expected accessibility at motif or kmer sites.

Use when after bias correction of ATAC-seq reads (via ATACorrect) when you have a bias-corrected bigWig file and need to measure transcription factor footprint strength within defined accessible regions (peaks, motif sites, or called footprint boundaries).

Use when you have a cooler file (.cool or .mcool) from a Hi-C or micro-C experiment and need to quantify the total number of sequencing reads assigned to each genomic bin to assess coverage uniformity, identify poorly-sequenced regions, or prepare bin-level weights for downstream normalization.

Use when when setting up a bioinformatics pipeline (particularly Hi-C data processing) that depends on multiple external binaries with version constraints, and you need to configure the environment in a way that is both portable across systems and reproducible across runs.

Use when when setting up a bioinformatics pipeline (such as HiC-Pro) that depends on multiple compiled or independently distributed binaries and you need to confirm that all required tools are installed, executable, meet version requirements (e.g., samtools ≥1.

Use when before executing a complex bioinformatics pipeline (such as Hi-C data processing) that depends on multiple third-party tools with explicit version constraints.

Use when you have raw methylation call files from Bismark, MethylDackel, or similar bisulfite alignment tools (bedGraph, cytosine report, or tabix-indexed formats) and need to import them into R as methylRaw or methylRawListDB objects for downstream differential methylation analysis, quality.

Use when when you have computed z-score deviations for genomic annotations (e.g., motifs) across multiple cells or samples and need to quantify uncertainty in their variability rankings before performing differential or comparative analyses.

Use when when you have computed per-bin insulation scores from a Hi-C cooler file using cooltools.insulation and need to identify discrete genomic boundaries that separate topological domains.

Use when you have executed a complex multi-step processing pipeline (e.g., ENCODE Hi-C uniform processing pipeline) and need to confirm that the generated output files match a known reference baseline.

Use when when you have aligned ChIP-Seq reads (single-end BED or paired-end BEDPE format) and need to identify enriched genomic regions by comparing ChIP signal against control background, with the ability to customize fragment length estimation, local bias calculation, and peak score thresholds.

Use when when beginning peak calling on ChIP-Seq data: you have raw single-end or paired-end BED/BEDPE alignment files for both ChIP and control samples and need to remove duplicate reads before predicting fragment length and building local bias models.

Use when after duplicate filtering and fragment length prediction (d) in ChIP-Seq analysis, when you need to convert discrete read alignments into continuous coverage signal for comparison against control background.

Use when you have loaded raw ATAC-seq fragment counts into a SummarizedExperiment object and are preparing to compute motif deviations.

Use when you have ATAC-seq BAM files aligned to a reference genome, a set of transcription factor motif locations (BED format), and you need to determine which motifs are actually occupied by proteins in your cell type or condition.

Use when when you have filtered ATAC-seq or DNAse-seq peak counts (after GC bias correction, sample filtering, and peak filtering) and wish to measure how strongly each annotation (motif or kmer) influences chromatin accessibility variability in each sample relative to a background expectation.

Use when use this skill after performing Tn5 bias correction and footprint scoring on ATAC-seq BAM files when you need to inspect the spatial distribution of Tn5 insertions around transcription factor binding sites, validate footprinting quality, or communicate differential TF occupancy patterns.

Use when after you have (1) corrected ATAC-seq BAM files for Tn5 insertion bias using ATACorrect, (2) computed per-base footprint scores using ScoreBigwig, (3) obtained a motif database (e.

Use when you have a backed AnnData object populated with fragment coordinates (stored in .obsm['fragment_paired'] or .

Use when you have sparse, single-cell or bulk ATAC/DNAse-seq data from multiple cell types or conditions (e.g., GM vs H1 cell lines), pre-filtered and GC-bias-corrected, with motif-to-peak matches already computed.

Use when you have generated a .hic contact map from Hi-C raw sequencing data and need to identify topologically associating domains (TADs) or other chromatin structural boundaries.

Use when when you have loaded a cooler file containing Hi-C contact matrices and need to quantify how contact probability decays with genomic distance within a single chromosome.

Use when you have executed an end-to-end SnapATAC2 pipeline on the pbmc10k_multiome dataset (or a similar single-cell ATAC-seq dataset with a published reference) and need to validate that spectral embedding, Leiden clustering, and UMAP layout have converged to expected cluster identities and.

Use when when you need to reproduce or validate benchmark comparisons between clustering methods on single-cell chromatin accessibility data, particularly when the source publication reports multiple accuracy metrics across heterogeneous datasets and you must decide which method variant (e.