Biomarker discovery and AI-assisted prediction continued to focus on practical decision points in oncology. A new study proposed a simple blood formula to predict survival in colorectal cancer patients with lung-spreading disease, aiming to help stratify outcomes without adding expensive testing. Alongside that, researchers developed an AI approach combining a slide-chart workflow to predict colorectal cancer risk, reflecting the industry’s push to convert complex clinical patterns into actionable risk scoring. Separately, a broader analysis in cancer genetics linked RNA splicing errors in drug transporter genes to survival across 33 tumor types—suggesting that transcript-level regulatory defects may become a new layer of prognostic biomarker strategy. Overall, the combined thrust is toward lower-friction tools—blood-based and model-driven—to support earlier therapeutic decisions while integrating molecular signals that go beyond single-gene mutations.