期间，我协助导师完成细胞转染，接触了质粒构建与克隆、PCR扩增、BCA蛋白定量、Western blot等实验技术，通过油镜观察对比了细胞内胆固醇的含量差异。这些内容与我在学校物化生实验室打下的基础成功衔接，让我更快进入了科研实操的状态。

During the program, I assisted my mentor with cell transfection and gained hands-on experience with experimental techniques including plasmid construction and cloning, PCR amplification, BCA protein quantification, and Western blotting. I also used an oil-immersion microscope to observe and compare differences in intracellular cholesterol levels. These techniques built naturally on the foundation I had developed in my school’s physics, chemistry, and biology laboratories, allowing me to adapt more quickly to hands-on scientific research.

Western blot实验让我印象最深。这次实验中，Z variant突变体样本的离心沉淀组分出现了清晰条带，野生型WT对照却无明显条带。这些实验让原先停留在理论层面的知识变得具体，也让我对蛋白突变与聚集机制有了更直观的理解。

The Western blot experiment left the strongest impression on me. In this experiment, a clear band appeared in the centrifuged pellet fraction of the Z-variant mutant sample, while no obvious band was observed in the wild-type (WT) control. Seeing these results transformed concepts that I had previously encountered only in theory into something tangible and gave me a much more intuitive understanding of protein mutations and aggregation mechanisms.

AISSU物化生实验室里养成的动手习惯和求证意识，让我面对一次次不如预期的结果时，不再急于否定自己，而是先回到实验记录中逐项排查。这对我实验期间帮助很大，让我在反复失败中依然能保持耐心，保持好奇，在一次次实践中不断靠近问题本质。

The hands-on habits and evidence-based mindset I had developed in AISSU’s science laboratories also changed the way I responded when experiments did not produce the results I expected. Instead of immediately doubting myself, I learned to return to my experimental records and troubleshoot each step systematically. This helped me enormously throughout the program. Even after repeated setbacks, I was able to remain patient and curious, gradually getting closer to the root of each problem through continuous experimentation.

在质粒提取实验中，我曾遇到样品浓度达标但纯度不达标的问题。逐步复盘后，我意识到，科研不能简单照搬流程，每个步骤背后都有原理，只有吃透底层逻辑，才能精准排查问题。

During a plasmid extraction experiment, I encountered a sample whose concentration met the required standard but whose purity did not. As I reviewed the procedure step by step, I realized that scientific research is not simply about following a protocol. Every step has an underlying scientific rationale, and only by understanding that logic can we accurately identify the source of a problem.

干实验建模的经历进一步重塑了我的科研思维。初期我尝试了特征筛选、多种模型架构和反复验证，最终发现初始模型预测性存在明显局限。于是我摒弃“追求高相关性”的固有思维，选择了更简洁、更稳定、更具有说服力的Ridge基线模型。因此，我明白了模型评估不能只看训练集效果，有效的验证与稳定的数据表现，远比单纯的高相关性更重要。

My experience with computational modeling further reshaped the way I approached research. At first, I experimented with feature selection, multiple model architectures, and repeated validation, only to discover clear limitations in the predictive performance of my initial models. I therefore moved away from simply pursuing high correlations and adopted a simpler, more stable, and more convincing Ridge regression baseline model. Through this process, I learned that a model cannot be evaluated solely by how well it performs on the training data. Rigorous validation and stable performance matter far more than achieving an impressive correlation alone.

在借助Codex训练AI搭建预测模型时，我看清了AI的局限：它偶尔输出偏差结果，我无法**时间发现漏洞，走了不少弯路。我终于明白校长所说的“要善于利用AI，让AI成为我们的辅助工具”的真正意思。AI是辅助，而非标准答案，只有依托自身专业知识甄别结果、合理借力，才能让AI真正服务于科研。

While using Codex as an AI-assisted tool to help build predictive models, I also came to recognize the limitations of AI. At times, it produced inaccurate or misleading outputs, and because I could not always identify the problems immediately, I took several unnecessary detours. This experience helped me truly understand what our principal meant by “learning to use AI effectively and making it a tool that assists us.” AI should be an assistant, not a source of unquestionable answers. Only when we have enough knowledge to evaluate its outputs critically and use it appropriately can AI genuinely contribute to scientific research.

回望暑期，无数个“**次”——独立分析Western blot、搭建预测模型、认清AI的局限性、直面实验结果与预测不匹配的常态——拼凑出我的科研成长轨迹。

Looking back on the summer, countless “firsts” marked my growth as a young researcher: independently analyzing Western blot results, building predictive models, recognizing the limitations of AI, and learning to accept the fact that experimental results do not always match predictions.

感谢深圳湾实验室的平台与AISSU的支持，让我摆脱“跟随实验”的被动，能够全程自主规划、操作、复盘，完成“科研新手”的蜕变。

I am grateful to the Shenzhen Bay Laboratory for providing this research opportunity and to AISSU for its support. This experience allowed me to move beyond simply following instructions and begin planning, conducting, and reviewing my own work with greater independence—an important step in my transformation from a research novice into a more self-directed young researcher.

任何值得抵达的地方，都没有捷径。

There are no shortcuts to any destination truly worth reaching.

精准的判断、熟练的能力、成熟的思维，都源于日复一日的打磨与复盘。未来，我将带着这份严谨与热爱，在更广阔的人生路上稳步前行。

Sound judgment, technical proficiency, and mature thinking are all built through persistent practice and reflection. In the future, I will carry this rigor and passion with me as I continue to explore a wider world and move steadily forward.