Differential Expression · Pipelines · Network Analysis
No Code & Agentic AI for Life Sciences · Session 04 · June 2026
A Count Matrix: Rows (Genes) x Columns (Samples). This is the starting point for statistical analysis.
DESeq2 is the industry standard for identifying Differentially Expressed Genes (DEGs).
log2FoldChange: Magnitude of change.
pvalue: Statistical significance.
padj: Adjusted p-value (FDR) — use this for filtering!
[Volcano Plot Schematic]
X-axis: Log2 Fold Change | Y-axis: -Log10 P-value
Highly significant, highly upregulated genes. Your primary candidates.
Highly significant, highly downregulated genes.
The "most significant" gene (highest point) isn't always the "most changed" gene (furthest right/left).
Input your list of DEGs. Enrichr checks for over-representation in:
Visualizes Protein-Protein Interaction (PPI) networks.
Agents excel at connecting raw gene lists to biological narratives.
Claude can search UniProt/PubMed for your STRING hub genes to see if they are known drug targets or prognostic markers.
Don't trust Claude's fold-change numbers if you haven't provided them. Always ground the interpretation in your actual DEG table data.
Next Session: Deep Learning Prediction Methods — Biomarker Classification and Model Selection.
© 2026 Md. Jubayer Hossain · No Code & Agentic AI for Life Sciences — Session 04