Bioinformatics Analysis

Overview

Computational biology and genomics analysis pipelines. GENERAL: not locked to any specific tool — use Scanpy, Seurat, DESeq2, or any appropriate package.

Common Workflows

RNA-seq Analysis

1. Quality control (FastQC, MultiQC)
2. Alignment (STAR, HISAT2) or pseudo-alignment (Salmon, kallisto)
3. Quantification (featureCounts, Salmon quant)
4. Normalization (DESeq2 vst/rlog, edgeR TMM)
5. Differential expression (DESeq2, edgeR, limma-voom)
6. Visualization (volcano plot, MA plot, heatmap)
7. Pathway analysis (clusterProfiler GO/KEGG, GSEA)

scRNA-seq Analysis (Scanpy / Seurat)

1. Quality control (filter cells: mito%, gene count, UMI count)
2. Normalization (library size, log1p)
3. Feature selection (highly variable genes)
4. Dimensionality reduction (PCA → UMAP/tSNE)
5. Clustering (Leiden, Louvain)
6. Marker identification (Wilcoxon, t-test)
7. Cell type annotation (manual markers or automated)
8. Trajectory analysis (PAGA, diffusion pseudotime)

GWAS / Genomics

1. Quality control (MAF, HWE, call rate, relatedness)
2. Association testing (PLINK, REGENIE)
3. Multiple testing correction (Bonferroni, FDR)
4. Manhattan plot, Q-Q plot
5. Fine-mapping, colocalization

Key Databases

• GEO (Gene Expression Omnibus): public expression datasets
• TCGA (The Cancer Genome Atlas): cancer genomics
• GTEx (Genotype-Tissue Expression): tissue-specific expression
• ClinVar: clinical variant interpretations
• UniProt: protein sequences and annotations
• Ensembl: genome annotation

File Formats

Format	Content	Tools
FASTQ	Raw sequencing reads	FastQC, Trimmomatic
BAM/SAM	Aligned reads	samtools, IGV
VCF	Variant calls	bcftools, GATK
h5ad	AnnData (scRNA-seq)	Scanpy
RDS	R object	Seurat, DESeq2
BED	Genomic regions	bedtools