A new bioinformatics approach for prediction of potential tumor neoantigens based on the cancer genome atlas dataset / 生物工程学报

Tumor-specific gene mutations might generate suitable neoepitopes for cancer immunotherapy that are highly immunogenic and absent in normal tissues. The high heterogeneity of the tumor genome poses a big challenge for precision cancer immunotherapy. Mutations characteristic of each tumor can help to distinguish it from other tumors. Based on these mutations' characteristic, it is possible to develop immunotherapeutic strategies for specific tumors. In this study, a tumor neoantigen prediction scheme was proposed, in which both the intracellular antigen presentation process and the ability to bind with extracellular MHC molecule were taken into consideration. The overall design is meritorious and may help reduce the cost for validation experiments compared with conventional methods. This strategy was tested with several cancer genome datasets in the TCGA database, and a number of potential tumor neoantigens were predicted for each dataset. These predicted neoantigens showed tumor type specificity and were found in 20% to 70% of cancer patients. This scheme might prove useful clinically in future.

Research progress of feature selection and machine learning methods for mass spectrometry-based protein biomarker discovery / 生物工程学报

With the development of mass spectrometry technologies and bioinformatics analysis algorithms, disease research-driven human proteome project (HPP) is advancing rapidly. Protein biomarkers play critical roles in clinical applications and the biomarker discovery strategies and methods have become one of research hotspots. Feature selection and machine learning methods have good effects on solving the "dimensionality" and "sparsity" problems of proteomics data, which have been widely used in the discovery of protein biomarkers. Here, we systematically review the strategy of protein biomarker discovery and the frequently-used machine learning methods. Also, the review illustrates the prospects and limitations of deep learning in this field. It is aimed at providing a valuable reference for corresponding researchers.