纳通道内λ-DNA过孔信号的小波去噪及统计分析

Wavelet denoising and statistical analysis of signal of λ-DNA translocation through nanopores

针对λ噬菌体中的脱氧核糖核酸(λ-DNA)通过纳米通道时的过孔信号噪音大,强度弱且非平稳的缺点,提出了使用具有良好时频域分辨能力的小波分析方法对其进行去噪处理,并对有效去噪后的λ-DNA过孔信号进行了统计分析.首先根据小波去噪原理,选择合适的小波函数,确定最佳的分解层数并选取合适的阈值,对实验采集到的含噪声信号进行去噪处理.根据最终去噪效果可得,以sym7为小波基函数、分解层数5层、使用默认软阈值可以有效降低信号中的噪声,提高信噪比.然后,对具有48 000个碱基对(48 kbp)的λ-DNA通过60 nm氮化硅(SiN)纳米孔的特征信号进行了统计分析,分析结果表明,阻塞电流和过孔时间分别符合双峰高斯分布和偏正态分布,这为后续DNA分子的辨识工作提供了依据.

To overcome the shortcomings that the signal of λ-DNA translocation through nanopores has high-noise, weak-strength and non-stationary, a wavelet analysis method with excellent resolution capabilities in time and frequency domain is proposed to analyze the signal of λ-DNA translocation. A statistical analysis is applied to the λ-DNA signal after effective denoising. First, according to the principle, suitable wavelet base function and the best decomposition level are chosen, and the denoising threshold is determined. The results indicate that by using sym7 wavelet base function, decomposition level at 5 and using default soft threshold the noise in the signal can be reduced effectively and the signal to noise ratio （SNR）can be improved. And then, the characteristic current signals, which arise from λ-DNA molecules of 48 kbp translocating through silicon nitride nanopores of 60 nm are statistically analyzed. The analysis results show that the current blockade and translocation time correspond to bimodal and skewed normal distribution respectively, which provides basis for subsequent DNA molecular identification.