Data analysis, machine learning, LLM apps, and prompt design (9.1K skills)
Filter and rank variants from exome or genome sequencing for rare disease analysis. Combine population frequency, clinical annotations, predicted effects, and inheritance patterns to produce prioritized candidate lists and evidence-based tiers for further review.
Create sample-specific FASTA sequences by applying recorded genetic variants to a reference with bcftools consensus. Select haplotypes, representing alternative inherited sequences, mask low-coverage regions, and export coordinate mappings between the reference and resulting sequence.
Read and write compressed sequence files with Biopython, including FASTA and FASTQ. Convert between compression formats, count sequence records, and use BGZF, an indexable compression format, to retrieve individual records by ID.
Analyze CRISPR base editing and prime editing results from sequencing data for a target region. Use CRISPResso2 to quantify intended changes, unintended nearby edits, and DNA insertions or deletions, and compare editing outcomes.
Check flow cytometry and CyTOF data quality before analysis. Use R examples to detect unstable flow rates, changing signals, measurements at detector limits, and dead cells, then produce cleaned data, plots, and sample summaries.
Load and inspect FCS files containing flow or mass cytometry measurements of cells. Explore measurement channels and sample metadata, extract or combine data, create basic plots, and save files for subsequent analysis.
Add gene names and biological context to CNVs, regions where DNA copy numbers change. Cross-reference cancer genes and ClinVar records, analyze affected pathways, and produce tables that support interpretation of copy number results.
Calculate tumor mutational burden, the number of protein-altering mutations per megabase, from tumor sequencing files. Apply quality filters, normalize by the tested region, classify results using selected thresholds, and summarize multiple samples.
Visualize copy number variation, changes in the number of DNA copies, using genome-wide plots, chromosome views, and sample heatmaps. Turn CNVkit or GATK results into figures for inspection, comparison, and publication.
Retrieve variant frequencies from gnomAD to support rare disease analysis. Check how often a genetic change appears overall or within ancestry groups, compare exome and genome records, and filter candidate variants using a chosen frequency threshold.
Look up genetic variants by rsID, a variant identifier, or genomic position. Retrieve dbSNP details and related ClinVar and gnomAD annotations, with examples for batch searches and extracting key fields.
Turn ChIP-seq signals and peaks into heatmaps, average signal plots, and genome browser tracks. Use deepTools and R tools to compare samples and examine patterns around gene starts, peaks, or selected genomic regions.
Annotate ChIP-seq peaks with ChIPseeker to connect detected regions to nearby genes and genomic features. Export annotation tables, plot distances from transcription start sites, compare experiments, and examine biological processes associated with peak-linked genes.
Retrieve clinical classifications, population frequencies, and other genetic variant annotations from multiple databases through myvariant.info. Query individual variants or batches, select relevant fields, and organize results into tables for further analysis.
Identify enriched regions in aligned ChIP-seq data using MACS3 or MACS2. Choose narrow or broad peaks, compare against input controls, adjust fragment settings, and export peak coordinates or signal tracks for visualization.
Assess ChIP-seq experiment quality before further analysis. Calculate enrichment, read duplication, and agreement between replicate experiments, then review quality metrics and plots to understand sample quality and identify potential problems.
Compare ChIP-seq peaks, genomic regions enriched for protein binding, across experimental conditions with DiffBind. Use replicate samples to calculate binding changes and statistical significance, visualize sample relationships, and export results as CSV or BED files.
Analyze ChIP-seq or ATAC-seq peak regions for DNA motifs, recurring sequence patterns. Use HOMER and MEME-ChIP to discover motifs, test known motif enrichment, locate matches, and summarize results with sequence logos.
Look up how genes and genetic variants relate to drug response using PharmGKB annotations and CPIC dosing guidelines. Includes Python examples for classifying drug metabolism from allele pairs and collecting drug–gene information into tables.
Narrow genetic association signals from GWAS, a genome-wide association study, to likely causal variants using SuSiE or FINEMAP. Estimate probabilities for individual variants and build candidate sets to prioritize functional follow-up experiments.
Check Mendelian randomization results for bias from genetic variants that affect an outcome through other pathways. Use outlier removal, regression diagnostics, and direction checks to compare corrected estimates and assess how consistently different methods support the findings.
Calculate polygenic risk scores, which combine effects from many genetic variants, using GWAS summary statistics and individual genotypes. The skill covers PRSice-2, LDpred2, PRS-CS, score normalization, population percentiles, and predictive performance evaluation.
Analyze whether gene expression, DNA methylation, or other molecular measurements carry the effect of a genetic variant to an outcome. Estimate direct and indirect effects, their uncertainty, and the proportion explained by the intermediate measurement.
Read, write, and convert multiple sequence alignments, files containing biological sequences arranged for comparison, using Biopython. Convert between formats such as Clustal, FASTA, and PHYLIP, extract selected sequences or regions, and process files in batches.