Design and Implementation of Network Transfer Protocol for Big Genomic Data
Design and Implementation of Network Transfer Protocol for Big Genomic Data
复制标题
基因组大数据网络传输协议的设计与实现
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
F. Saeed
中科院分区：
文献类型：
--
作者：
Mohammed Aledhari;F. Saeed
由于下一代测序技术 (NGS) 及其在短时间内生成大量数据的能力，基因组数据呈指数级增长。 NGS 技术生成大量基因组数据，需要在不同位置之间高效、可靠地交换。当前的网络传输协议依赖于传输控制协议（TCP）或用户数据报协议（UDP）协议，忽略数据大小和类型。 HTTP 等通用应用层协议是为各种数据类型而设计的，对于基因组数据来说并不是特别有效。因此，我们提出了一种新的基因组数据数据感知传输协议，可以提高网络吞吐量并减少延迟，称为基因组文本传输协议（GTTP）。在本文中，我们为大型基因组 DNA 数据集设计并实现了一种新的网络传输协议，该协议依赖于超文本传输协议 (HTTP)。对 HTTP 内容编码的修改已经完成，可以使用机器对机器 (M2M) 和客户端-服务器拓扑传输大型基因组 DNA 数据集。我们的结果表明，我们对 HTTP 的修改将传输的数据减少了原始数据的 75%，并且仍然能够在客户端重新生成数据以进行生物信息学分析。因此，与常规 HTTP 相比，使用 GTTP 传输数据要快得多（大约比 HTTP 快 8 倍）。
Genomic data is growing exponentially due to next generation sequencing technologies (NGS) and their ability to produce massive amounts of data in a short time. NGS technologies generate big genomic data that needs to be exchanged between different locations efficiently and reliably. The current network transfer protocols rely on Transmission Control Protocol (TCP) or User Data gram Protocol (UDP) protocols, ignoring data size and type. Universal application layer protocols such as HTTP are designed for wide variety of data types and are not particularly efficient for genomic data. Therefore, we present a new data-aware transfer protocol for genomic-data that increases network throughput and reduces latency, called Genomic Text Transfer Protocol (GTTP). In this paper, we design and implement a new network transfer protocol for big genomic DNA dataset that relies on the Hypertext Transfer Protocol (HTTP). Modification to content-encoding of HTTP has been done that would transfer big genomic DNA datasets using machine-to-machine (M2M) and client(s)-server topologies. Our results show that our modification to HTTP reduces the transmitted data by 75% of original data and still be able to regenerate the data at the client side for bioinformatics analysis. Consequently, the transfer of data using GTTP is shown to be much faster (about 8 times faster than HTTP) when compared with regular HTTP.