女性最常见的癌症类型是乳腺癌，仅次于肺癌。本文总结了基因组学和表观遗传学的变化以及逐步增加的生物活动。肿瘤通过涉及单独异常基因的一系列阶段来发展。尽管许多疾病会导致DNA突变，但大多数治疗方法旨在缓解症状而不是改变DNA。聚类短回文重复序列（CRISPR）或Cas9是发现和确认肿瘤发生性基因靶标的主要方法。开发了一种具有表达编程模型的Kohonen神经网络用于基因选择。遗传选择中的主要问题是减少选择的特征数量同时保持准确性。这一目的是系统地实现的。最终，该方法优于现有的量子松鼠启发式算法和循环神经网络对阻性查询算法在基因选择方面的表现。 KNNet-EPM模型采用了一种表达编程方法来识别乳腺癌基因生物标志物。该方法的RAE为42%，敏感性为93%，f1得分为88%，准确率为98%，kappa得分为83%，特异性为92%，MAE为30%。© 2023. 作者，专属许可给 Springer-Verlag GmbH Germany，属于 Springer Nature 的一部分。

Women's most frequent type of cancer is breast cancer, second only to lung cancer. This paper summarizes changes in genomics and epigenetics and incremental biological activities. A tumour develops through a series of phases involving a separate abnormal gene. Even though many diseases cause DNA mutations, most treatments are designed to relieve symptoms rather than change the DNA. Clustering short palindromic repeats (CRISPR) or Cas9 is the primary approach for discovering and confirming tumorigenic genomic targets. A Kohonen neural network with an expression programming model was developed for gene selection. The main problem in genetic selection is reducing the number of features chosen while maintaining accuracy. This purpose is accomplished systematically. In the end, the approach method performed better than the existing quantum squirrel-inspired algorithm and the recurrent neural network oppositional call search algorithm for genetic selection. The KNNet-EPM model used an expression programming approach to identify gene biomarkers for breast cancer. This method was achieved with RAE of 42%, sensitivity of 93%, f1 score of 88%, accuracy of 98%, kappa score of 83%, specificity of 92% and MAE of 30%.© 2023. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.