MOTIVATION: DNA copy number variations (CNVs) exert a profound impact on major genetic disorders in humans. Although multiple sequencing technologies have become the first line of molecular diagnosis for CNVs, existing tools are unable to resolve the pathogenicity of CNVs directly from raw sequencing data. RESULTS: We developed CNVSeeker, a one-stop and easy-to-use pipeline that provides comprehensive analysis from raw sequencing data to variant interpretation reports, and supports multiple types of sequencing data including short-read data such as whole genome sequencing data and whole exome sequencing data, and long-read sequencing data from Pacific Biosciences HiFi platform or Oxford Nanopore Technologies platform. Through extensive benchmarking, CNVSeeker demonstrated comparable enhancement over the state-of-the-art methods for CNV calling. Moreover, CNVSeeker enables significantly precise variant classification with an accuracy of ∼87%. By applying CNVSeeker to 1946 individuals with autism spectrum disorder (ASD), a total of 133 ASD-associated CNVs in 122 patients were identified, yielding a diagnostic yield of ∼6.3%. Additionally, we have also provided a user-friendly webserver for intuitive visualization of results. This study highlights the potential of CNVSeeker to benefit clinicians and geneticists with limited bioinformatic skill by aiding them interpret CNVs directly from various types of raw sequencing data for auxiliary disease diagnosis. AVAILABILITY AND IMPLEMENTATION: The web server is freely available at https://genemed.tech/cnvseeker and the open-source code can be found at https://github.com/lovelycatZ/CNVSeeker.

With the increasing popularity of services published online, energy consumption of services computing systems is growing dramatically. Besides Quality of Service (QoS), energy efficiency has become an important issue and drawn significant attention. However, energy efficient request scheduling and service management for large-scale services computing systems face challenges because of the high dynamics and unpredictability of request arrivals. In this paper, we jointly consider the conflicting metrics of performance, queue congestion and energy consumption. We propose a distributed online scheduling and management algorithm which does not require any priori statistical knowledge of request arrivals. Mathematical analysis is conducted which demonstrates that our algorithm can achieve arbitrary tradeoff between performance and energy efficiency. Numerical and real trace data based experiments are carried out to validate the effectiveness of our algorithm in optimizing energy efficiency while stabilizing the system.

Mobile cloud computing (MC2) is emerging as a promising computing paradigm which helps alleviate the conflict between resource-constrained mobile devices and resource-consuming mobile applications through computation offloading. In this paper, we analyze the computation offloading problem in cloudlet-based mobile cloud computing. Different from most of the previous works which are either from the perspective of a single user or under the setting of a single wireless access point (AP), we research the computation offloading strategy of multiple users via multiple wireless APs. With the widespread deployment of WLAN, offloading via multiple wireless APs will obtain extensive application. Taking energy consumption and delay (including computing and transmission delay) into account, we present a game-theoretic analysis of the computation offloading problem while mimicking the selfish nature of the individuals. In the case of homogeneous mobile users, conditions of Nash equilibrium are analyzed, and an algorithm that admits a Nash equilibrium is proposed. For heterogeneous users, we prove the existence of Nash equilibrium by introducing the definition of exact potential game and design a distributed computation offloading algorithm to help mobile users choose proper offloading strategies. Numerical extensive simulations have been conducted and results demonstrate that the proposed algorithm can achieve desired system performance.

With the increasing popularity of web services on the Internet, service selection has become an important issue in large-scale web service systems. In recent years, with the growing energy consumption associated with IT systems and services, energy efficiency has drawn extensive attention in service allocation and selection. However, existing approaches for energy efficient web service selection face great challenges, because of the high dynamics and unpredictability of task arrivals, and large scale of current web service systems. In this paper, we present a distributed online approach for web service selection that jointly considers the conflicting performance metrics including response time, queue congestion and energy consumption. Targeting at optimizing the long-term average system reward, our approach does not require any priori knowledge of the statistics or prediction on task arrivals. Mathematical analysis as well as simulation experiments demonstrate its effectiveness in optimizing energy efficiency while stabilizing the system.