Claude Skills by Pavel-Kravchenko

Biochemistry fundamentals for computational biologists — Beer-Lambert law, spectrophotometric assays, Michaelis-Menten kinetics, linearization methods, and inhibition models

Cancer transcriptomics — melanoma subtype classification (Tirosh/Harbst), preprocessing pipeline, PCA/t-SNE, hierarchical clustering, random forest, and Kaplan-Meier survival analysis

Integrative bioinformatics pipeline — sequence QC, BLAST identification, MSA, phylogenetics, structure analysis, GO enrichment, and publication figures

scRNA-seq cell type annotation — manual marker scoring, SingleR reference-based, and CellTypist automated classification

ChIP-seq pipeline — QC, alignment, deduplication, peak calling with MACS2, and signal normalization with deepTools

CITE-seq and Multiome integration — ADT normalization (CLR/DSB), WNN graph construction, and paired RNA+ATAC analysis with muon

Clinical genomics — ACMG/AMP variant classification, ClinVar queries, and clinical reporting workflows

DNA copy number analysis — read depth normalization, CBS segmentation, CN state calling, and genome-wide visualization

Multi-omics data harmonization — normalization strategies, missing data imputation, batch correction, and integration approaches (MOFA2, DIABLO)

PyTorch deep learning for biological sequences — when to use DL vs classical ML, CNN architecture for motif detection, one-hot encoding, and the standard training loop

Differential binding analysis for ChIP-seq: DiffBind workflow, consensus peaks, normalization, and MA/volcano plots. Use when comparing ChIP-seq signal between conditions.

scRNA-seq dimensionality reduction and clustering: PCA, k-NN graph, UMAP, Leiden. Parameter selection guide, implementation patterns, and pitfalls.

Differentially Methylated Regions (DMRs)

Molecular docking: ligand preparation, receptor setup, AutoDock Vina workflow, scoring functions, and binding pose analysis. Use when predicting protein-ligand interactions.

Enzyme kinetics computational patterns — fitting Michaelis-Menten with scipy, bootstrap confidence intervals, inhibition type determination, allosteric cooperativity, and multi-substrate kinetics

Epigenetic Clocks and Aging Analysis with Matplotlib

Functional Annotation of Metagenomes with NumPy

GRN inference methods: correlation, mutual information (ARACNE), and random forest (GENIE3). Decision table for method selection, evaluation patterns, and key pitfalls.

In silico restriction digestion, compatible end detection, primer design (Tm models), and gel simulation

Genome assembly algorithms (OLC and de Bruijn graph), k-mer selection, assembler comparison table, and SPAdes/Flye/hifiasm usage.

Genome-Wide Association Studies (GWAS) with NumPy

HLA Typing and Antigen Presentation

Immune repertoire sequencing — TRUST4/MiXCR workflows, diversity metrics, clonal tracking, Morisita-Horn overlap, VDJdb lookup, CDR3 Hamming clustering.

Isoform analysis with long reads — Minimap2 splice alignment, bambu isoform discovery, DRIMSeq differential isoform usage.

LC-MS metabolomics data preprocessing: peak picking, retention time alignment, gap filling, and adduct detection. Use when processing raw mass spectrometry data for metabolomics studies.

Long Non-Coding RNA: Discovery and Classification

Machine learning for bioinformatics — feature engineering for sequences, promoter classification, train/test splits, logistic regression, random forest, and bio-specific pitfalls

CRISPR Screen Analysis with MAGeCK with MAGeCK

Flux balance analysis and metabolic modeling with COBRApy. Use when predicting metabolic fluxes, simulating gene knockouts, or analyzing stoichiometric models.

Metabolite identification from MS/MS spectra: spectral matching, molecular formula prediction, and database searching (HMDB, KEGG). Use when annotating unknown metabolites.

Microbial diversity analysis: alpha/beta diversity metrics, OTU/ASV methods, taxonomy assignment, and community comparison. Use when analyzing 16S amplicon or microbiome data.

miRNA-seq pipeline: adapter trimming, alignment to miRBase, quantification, DE analysis, and target prediction. Use when processing small RNA sequencing data.

mixOmics PLS-DA and DIABLO for supervised multi-omics integration and feature selection. Use when classifying samples or selecting biomarkers from multi-omics data.

MOFA2 unsupervised multi-omics factor analysis: variance decomposition, factor interpretation, and shared/view-specific signal separation. Use when integrating multiple omics layers.

Population genetics and molecular evolution — Hardy-Weinberg, Wright-Fisher drift, selection models, dN/dS, Tajima's D, Fst, and the neutral theory

Graph Neural Networks for Molecular Property Prediction with RDKit

Molecular Modeling with NumPy

Community detection in biological networks — Louvain/Leiden algorithms, modularity Q, WGCNA co-expression modules, and Cytoscape export

NGS platform comparison, FASTQ format, Phred quality scores, and QC metrics. Reference for sequencing technology selection and read quality assessment.

- Implement Lagrange and Newton interpolation polynomials and explain the Runge phenomenon - Apply cubic spline interpolation to reconstruct missing time points in biological time series - Compute num

ONT Data Processing with NumPy

Viral phylodynamics — molecular clocks, root-to-tip regression, time-scaled phylogenies, Bayesian skyline plots, and phylogeography tool selection

Population genetics — Hardy-Weinberg equilibrium, Wright-Fisher drift simulation, selection models, molecular clock, dN/dS, Tajima's D, Fst, and linkage disequilibrium

PPI network construction and analysis with NetworkX and STRING DB: centrality metrics, hub/bottleneck classification, scale-free properties, and community detection.

Core promoter elements: TATA box, CpG islands, PWM construction, and TFBS scanning. Companion reference card to bio-applied-regulatory-analysis.

Proteomics data analysis — peptide identification, quantification, PTM analysis, and protein inference workflows

16S rRNA amplicon analysis with QIIME2: DADA2 denoising, taxonomy assignment, alpha/beta diversity, and differential abundance. Use when analyzing 16S microbiome data.

QSAR modeling: molecular descriptors, fingerprints, random forest/SVM models, applicability domain, and model validation. Use when building structure-activity relationship models.

RDKit fundamentals: SMILES parsing, molecular properties, substructure search, fingerprints, and chemical similarity. Use when performing cheminformatics operations in Python.

Promoter and regulatory sequence analysis: TATA box detection, CpG island scanning, PWM/PFM construction, and TFBS scanning. Reference for computational promoter analysis.