To test a language model’s understanding of concept relationships in a specific scientific field (here, eukaryotic gene expression regulation), use this structured prompt. It evaluates the model’s ability to map interconnections, trace causal chains, analyze perturbations, and apply knowledge to hypothetical scenarios:

Prompt:

Your task is to demonstrate a deep understanding of the relationships between key scientific concepts in the field of Gene Expression Regulation in Eukaryotic Cells. Focus on connecting concepts (not just listing facts) and answer the following questions clearly:

1. Concept Interconnections:
   List 8–10 core concepts (e.g., histone acetylation, promoter, spliceosome) and explain how each interacts with at least two other concepts. Example: "Enhancers bind transcription factors; this interaction increases promoter accessibility to RNA polymerase II."

2. Causal Chain Walkthrough:
   Describe the sequence of events from an epigenetic modification (e.g., DNA methylation) to the synthesis of a functional protein. Highlight key concepts at each step and their dependencies (e.g., how methylation blocks transcription factor binding, reducing mRNA production).

3. Perturbation Impact:
   If a drug inhibits spliceosome activity, what immediate and long-term effects would this have on gene expression? Connect your answer to the concepts above, explaining how the perturbation disrupts normal relationships.

4. Hypothetical Mutation Analysis:
   A mutation damages the binding site of a transcription factor (TF) that activates a tumor suppressor gene. How would this mutation affect the gene’s expression, and what downstream biological outcomes might follow? Justify your answer using relevant concepts.

Your response should show logical, evidence-based connections between concepts to prove mastery of the field’s underlying relationships.

Why This Works:

• Specificity: Targets a well-defined scientific domain (eukaryotic gene regulation) with clear, interconnected concepts.
• Relationship Focus: Requires the model to link concepts (not just recall facts) via interconnections, causal chains, and perturbation effects.
• Application: Tests the model’s ability to apply knowledge to real-world-like scenarios (mutations, drug effects) — a key indicator of deep understanding.

This prompt will reveal if the model can synthesize information rather than just regurgitate lated facts.

作者声明:本文包含人工智能生成内容。