Hengjian Cui

Accurately predicting HIV-1 sensitivity to broadly neutralizing antibodies (bNAbs) is a critical step in advancing the development of effective therapeutic and preventive strategies against HIV-1. Due to the virus's extraordinary genetic diversity and rapid mutation rate, traditional experimental methods for assessing bNAb sensitivity are labor-intensive, time-consuming, and limited in scalability. However, current machine learning approaches prioritize feature significance while neglecting to address class imbalance and systematic biases in multi-laboratory datasets. To address these limitations, we develop the Statistical Distribution Sampling (SDS) method, a novel statistical framework that systematically addresses data imbalance and cross-institutional variance. Firstly, the amino acid sequence is converted into a numerical vector via the k-string method. Then, antibodies are classified as effective or ineffective based on IC50/IC80 thresholds, with a stratified random subset selected as the Sampling generator. Each vector dimension's distribution is characterized through histogram analysis and kernel density estimation. Afterwards, 500 data points are sampled from the distribution of each column for each category. The assembled Statistical Distribution Sampling (SDS) matrix integrates column features, which combined with the Sampling generator, produces the final training dataset. Antibody efficacy predictions are ultimately generated through integration with the Random Forest algorithm. This approach enables comprehensive prediction of all database sequences while achieving strong predictive performance. Through 10 independent trials per antibody, our method achieved statistically superior AUC and ACC performance compared to the LBUM and SLAPNAP benchmarks. Moreover, our method demonstrates significantly lower standard deviation across 10 runs than both counterparts, proving its consistent robustness across diverse antibody scenarios.

What is already known about this topic? Current HIV-1 subtype classification tools often rely on time-consuming alignment, whereas new non-alignment methods typically target single genes. China lacks a model for specifically predicting the non-B subtype strains prevalent domestically. What is added by this report? We developed a fast alignment-free method (SCCFV-RDA) for building multigene models. It achieved over 99.85% accuracy in classifying an international dataset and 99.7% accuracy in classifying Chinese pol gene data and showed superior recall for key circulating recombinant form subtypes. What are the implications for public health practice? This tool provides accurate and efficient computational support for the precise molecular surveillance of HIV-1 in China, thereby facilitating the formulation of targeted prevention and control strategies.

The genetic algorithm (GA) is a widely recognized optimization technique for addressing complex and NP-hard problems; however, its effectiveness is often hampered by issues such as premature convergence and population diversity. The selection operator, which determines the probability of individuals being chosen for reproduction, significantly influences its performance. This study presents a novel selection operator, which is specifically designed to strike a balance between selection pressure and the preservation of population diversity. At the first stage, we derived the mathematical properties and assessed the sampling accuracy using Pearson's chi-square test. For global investigation, we test its performance on benchmark TSPLIB instances and a real-world dataset of Pakistan's city coordinates in the context of the traveling salesman problem TSP. The results indicate that the newly proposed operator either outperforms or equals the performance of other operators across key metrics.