Yinhe Cao

Multiscale Analysis of Heart Rate Variability: A Comparison of Different Complexity Measures

Heart rate variability (HRV) is an important dynamical variable of the cardiovascular function. There have been numerous efforts to determine whether HRV dynamics are chaotic or random, and whether certain complexity measures are capable of distinguishing healthy subjects from patients with certain cardiac disease. In this study, we employ a new multiscale complexity measure, the scale-dependent Lyapunov exponent (SDLE), to characterize the relative importance of nonlinear, chaotic, and stochastic dynamics in HRV of healthy, congestive heart failure (CHF), and atrial fibrillation subjects. We show that while HRV data of all these three types are mostly stochastic, the stochasticity is different among the three groups. Furthermore, we show that for the purpose of distinguishing healthy subjects from patients with CHF, features derived from SDLE are more effective than other complexity measures such as the Hurst parameter, the sample entropy, and the multiscale entropy.

Protein Coding Sequence Identification by Simultaneously Characterizing the Periodic and Random Features of DNA Sequences

Most codon indices used today are based on highly biased nonrandom usage of codons in coding regions. The background of a coding or noncoding DNA sequence, however, is fairly random, and can be characterized as a random fractal. When a gene-finding algorithm incorporates multiple sources of information about coding regions, it becomes more successful. It is thus highly desirable to develop new and efficient codon indices by simultaneously characterizing the fractal and periodic features of a DNA sequence. In this paper, we describe a novel way of achieving this goal. The efficiency of the new codon index is evaluated by studying all of the 16 yeast chromosomes. In particular, we show that the method automatically and correctly identifies which of the three reading frames is the one that contains a gene.

Exploiting noise in array CGH data to improve detection of DNA copy number change

Developing effective methods for analyzing array-CGH data to detect chromosomal aberrations is very important for the diagnosis of pathogenesis of cancer and other diseases. Current analysis methods, being largely based on smoothing and/or segmentation, are not quite capable of detecting both the aberration regions and the boundary break points very accurately. Furthermore, when evaluating the accuracy of an algorithm for analyzing array-CGH data, it is commonly assumed that noise in the data follows normal distribution. A fundamental question is whether noise in array-CGH is indeed Gaussian, and if not, can one exploit the characteristics of noise to develop novel analysis methods that are capable of detecting accurately the aberration regions as well as the boundary break points simultaneously? By analyzing bacterial artificial chromosomes (BACs) arrays with an average 1 mb resolution, 19 k oligo arrays with the average probe spacing <100 kb and 385 k oligo arrays with the average probe spacing of about 6 kb, we show that when there are aberrations, noise in all three types of arrays is highly non-Gaussian and possesses long-range spatial correlations, and that such noise leads to worse performance of existing methods for detecting aberrations in array-CGH than the Gaussian noise case. We further develop a novel method, which has optimally exploited the character of the noise, and is capable of identifying both aberration regions as well as the boundary break points very accurately. Finally, we propose a new concept, posteriori signal-to-noise ratio (p-SNR), to assign certain confidence level to an aberration region and boundaries detected.

Multiscale analysis of heart rate variability

Biological time series are often highly nonlinear and nonstationary. To effectively characterize the complexity of biological signals, we propose a new multiscale analysis method. It has a distinguished feature of scale isolation, and thus can readily deal with nonstationarity in biological signals. By analyzing a number of heart rate variability data, we show that the method can accurately distinguish between healthy subjects and patients with congestive heart failure. Furthermore, our analysis suggests that the dimension of the dynamics of the cardiovascular system is lower under the healthy than under diseased conditions. This is compatible with the observation that a healthy cardiovascular system is a tightly coupled system with coherent functions, while components in a malfunctioning cardiovascular system are somewhat loosely coupled and function incoherently. Therefore, if cardiovascular dynamics could be deterministically chaoslike, it would be more likely to be detected in healthy subjects

Detection of gene copy number change in array CGH data

Developing effective methods for analyzing array-CGH data to detect chromosomal aberrations is very important for the diagnosis of pathogenesis of cancer and other diseases. Current analysis methods, being largely based on smoothing and/or segmentation, are not quite capable of detecting both the aberration regions and the boundary break points very accurately. This is undesirable, since each point in the array represents a gene. Furthermore, when evaluating the accuracy of an algorithm for analyzing array-CGH data, it is commonly assumed that noise in the data follows normal distribution. A fundamental question is whether noise in array-CGH is indeed Gaussian, and if not, can one exploit the characteristics of noise to develop novel analysis methods that are capable of detecting accurately the aberration regions as well as the boundary break points simultaneously? By analyzing bacterial artificial chromosomes (BACs) arrays, oligo-nucleotide arrays, and high density NimbleGen data, we show that when there are aberrations, noise in all three types of arrays is highly non-Gaussian and possesses long-range spatial correlations, and that such noise leads to worse performance of existing methods for detecting aberrations in array-CGH than the Gaussian noise case. We further develop a novel method, which has optimally exploited the characteristics of the noise, and is capable of identifying both aberration regions as well as the boundary break points very accurately

Distinguishing Healthy Subjects from Patients with Congestive Heart Failure Using Scale-Dependent Lyapunov Exponent

Heart rate variability (HRV) time series is highly nonlinear and nonstationary. To effectively characterize its complexity, we employ a newly developed multiscale complexity measure, the scale-dependent Lyapunov exponent (SDLE). We derive two readily computable features from the SDLE and show that they can readily distinguish healthy subjects from patients with congestive heart failure (CHF). The same task is evaluated using other complexity measures, including the Hurst parameter, the sample entropy, and the multiscale entropy. It is shown that for the purpose of distinguishing healthy subjects from patients with CHF, the features derived from the SDLE are much more effective than the Hurst parameter, the sample entropy, and the multiscale entropy.