Data analysis, machine learning, LLM apps, and prompt design (9.1K skills)
Explore which antigen fragments T-cell receptors may recognize. The skill covers ERGO-II setup, matching receptor sequences against known examples in VDJdb, grouping similar sequences, and summarizing candidate recognition patterns across a receptor collection.
Detect large DNA changes, such as deletions, insertions, and inversions, from aligned ONT or PacBio long reads. Use command-line tools to produce variant files, combine samples or callers, and filter or annotate results.
Identify genes whose main transcript form changes between conditions using IsoformSwitchAnalyzeR. Assess predicted effects on protein domains, protein production, and transcript degradation, then visualize individual switches and summarize patterns across the analyzed genes.
Use pVACtools to predict neoantigens, potential immune targets created by tumor mutations. Combine patient HLA types with annotated variants, then rank peptide candidates using predicted binding, mutation frequency, and available gene expression data.
Process PacBio Iso-Seq data to discover and quantify full-length transcript variants, called isoforms. Follow read cleanup, clustering, genome alignment, and quality checks, then classify known and novel variants and examine differences in isoform usage.
Label cell types in imaging mass cytometry images using napari, an image viewer, and marker overlays. Export labeled cells as training data, extend labels to similar cells, and generate image panels for annotation review.
Use medaka to correct errors in genome assemblies and call haploid variants from Oxford Nanopore reads. Select a matching neural network model, process existing alignments or raw reads, and filter variant results by quality.
Analyze metabolite measurements in R to find differences between experimental groups and visualize sample patterns. Use statistical tests, classification models, and performance measures to rank relevant metabolites and explore potential biomarkers, which are measurable indicators of biological conditions.
Analyze DNA methylation data in R with methylKit. Import Bismark files, filter and normalize read coverage, compare samples, and identify methylation differences at individual DNA sites or across fixed genomic regions.
Identify nuclei and whole cells in preprocessed IMC or MIBI tissue images using Cellpose or Mesmer. Extract marker intensities, cell shapes, and positions, inspect segmentation overlays, and save labeled images and tables for further single-cell analysis.
Analyze 16S or ITS amplicon data, sequences from amplified marker regions, through the QIIME2 command line. Import reads, remove noise, assign taxonomy, compare microbial diversity and abundance between groups, and retain a record of processing steps.
Identify TADs, genome regions with frequent internal contacts, from Hi-C data using cooltools or HiCExplorer. Calculate boundary scores, extract regions, compare boundaries across conditions, and export results as plots or genomic coordinate files.
Compare bacterial genomes and metagenomes to track strains, which are groups within a species. Use MASH, sourmash, fastANI, and inStrain to examine sample similarities and genetic variation for outbreak, transmission, or within-host studies.
Turn metagenomic profiles, which describe microbial composition, into figures using R or Python. Create stacked bars, heatmaps, and sample comparison plots, then examine diversity within samples and differences between groups.
Align Oxford Nanopore and PacBio sequencing reads to a reference genome with minimap2. Choose settings for each read type and produce sorted alignment files for downstream variant detection or analysis of sequencing coverage.
Align bisulfite sequencing reads to a reference genome with Bismark. Prepare genome indexes, handle single or paired reads, and generate BAM alignment files carrying methylation information for WGBS and RRBS analysis.
Analyze serial liquid biopsy measurements to track ctDNA, tumor-derived DNA in blood. Summarize tumor fraction trends, mutation clearance, and rises after the lowest measurement through plots and structured reports using specified response thresholds.
Calculate concentrations of selected metabolites from MRM/SRM measurements, which track specific molecular signals. Use calibration curves and internal reference compounds, apply dilution corrections, and assess measurement accuracy, variability, detection limits, and quantification limits.
Analyze lipid measurements from LC-MS, a method combining separation and mass measurement. Organize lipid classes and chain composition, compare abundance between conditions, visualize changes, and relate results to lipid pathways using lipidr and annotation resources.
Analyze DNA methylation in Oxford Nanopore data without bisulfite conversion. Use modkit to summarize modified-base information into per-site methylation frequencies, inspect sample statistics, and focus analysis on CpG islands or promoter regions.
Preprocess LC-MS metabolomics data with MS-DIAL, from detecting peaks and matching them across samples to assigning metabolite candidates. Prepare filtered measurement tables for further analysis in R or Python, or export data for MetaboAnalyst.
Prepare IMC and MIBI images for cell segmentation, the process of separating individual cells. Load MCD or TIFF data, remove unusually bright pixels, normalize channel signals, and correct signal leakage between channels.
Map measured metabolites to biological pathways and identify pathways represented more than expected. Use enrichment tests and pathway structure analysis to interpret results, display metabolite values on pathway diagrams, and export sorted findings for further review.
Identify microbial taxa whose abundance differs between experimental groups while accounting for data measured as relative proportions. Provides R workflows using ALDEx2, ANCOM-BC2, and MaAsLin2 to filter statistical results, handle additional variables, and visualize differences.