Abstract:

With the rapid development of high-throughput sequencing technologies， bioinformatics has entered the era of multi-omics big data. Python， featured by its flexibility and abundant toolbox ecosystem， has become an indispensable analytical language in this field. Despite its wide adoption， Python still faced bottlenecks in processing ultra-large-scale datasets， algorithm generalizability and model interpretability， which restricted its in-depth application in bioinformatics. This review systematically summarized the applications of Python in genomics， transcriptomics， proteomics and metabolomics， focusing on its functions in data preprocessing， analytical workflow construction and data visualization. It elaborated on Python-based machine learning methods （e.g.， random forest， support vector machine） and deep learning approaches （e.g.， neural networks， graph convolutional networks） in biomarker screening， disease prediction and drug research and development. Typical research cases were analyzed to clarify current challenges and potential optimization directions. This review aimed to provide a systematic reference for researchers who adopt Python for multi-omics data analysis， help address practical issues including large-scale data processing and model optimization， and facilitate the progress of biological data analysis technologies.

Key words: bioinformatics, Python, machine learning, neural networks, multi-omics