genomics

Genomics and Transcriptomics Analysis

When to Use This Skill

• When data contains gene expression measurements (RNA-seq, microarray)
• When analyzing differential gene expression
• When performing pathway or gene set enrichment analysis
• When interpreting genetic variants or mutations

Core Concepts

Gene Expression Data Types

RNA-seq counts:

• Raw read counts per gene
• Requires normalization (TPM, RPKM, DESeq2)
• Suitable for differential expression analysis

Microarray intensities:

• Probe fluorescence intensities
• Log-transformed, background-corrected
• Legacy platform, less common now

Single-cell RNA-seq:

• Expression per cell (not bulk tissue)
• High sparsity (many zeros)
• Specialized analysis methods
• ⚠️ For differential expression across conditions, aggregate to the sample level (pseudobulk) — cells are not independent replicates (see below)

Gene Nomenclature

Human genes:

• Official symbols: HUGO Gene Nomenclature Committee (HGNC)
• Example: TP53 (tumor protein p53)
• Italicized in publications

Mouse genes:

• Similar to human but capitalization differs
• Example: Tp53 (first letter capital, rest lowercase)

Protein names:

• Not italicized
• Example: p53 protein

Always verify gene symbols - aliases and outdated names are common.

Differential Expression Analysis

Workflow

import pandas as pd
import numpy as np
from scipy.stats import ttest_ind
from statsmodels.stats.multitest import multipletests

# Load expression data (genes × samples)
# Rows = genes, Columns = samples
expr_data = pd.read_csv("expression_data.csv", index_col=0)

# Define groups
group1_samples = ["Sample1", "Sample2", "Sample3"]
group2_samples = ["Sample4", "Sample5", "Sample6"]

results = []

for gene in expr_data.index:
    group1_expr = expr_data.loc[gene, group1_samples]
    group2_expr = expr_data.loc[gene, group2_samples]

    # T-test
    t_stat, p_value = ttest_ind(group1_expr, group2_expr)

    # Fold change
    mean1 = group1_expr.mean()
    mean2 = group2_expr.mean()
    log2fc = np.log2(mean1 / mean2) if mean2 > 0 else np.nan

    results.append({
        "gene": gene,
        "log2FC": log2fc,
        "p_value": p_value,
        "mean_group1": mean1,
        "mean_group2": mean2
    })

results_df = pd.DataFrame(results)

# Multiple testing correction
results_df["p_adj"] = multipletests(results_df["p_value"], method="fdr_bh")[1]

# Define significant genes
significant = results_df[
    (results_df["p_adj"] < 0.05) &
    (abs(results_df["log2FC"]) > 1)  # 2-fold change
]

print(f"Significant genes: {len(significant)}")
print(f"Upregulated: {sum(significant['log2FC'] > 0)}")
print(f"Downregulated: {sum(significant['log2FC'] < 0)}")