局部序列比对算法及其并行加速研究进展

随着新一代测序技术的发展,传统的序列比对工具已无法满足测序产生的海量生物学数据分析处理的需求,研究如何利用最新的计算技术加速序列比对过程具有十分重要的意义。本文回顾了常用的局部序列比对算法,介绍了基于并行计算原理的序列比对算法的加速优化策略和主要进展,详细说明了如何利用最新的图形处理器（GPU）计算技术实现高性能的BLAST（basic local alignment search tool）比对算法。最后,结合实际需求,提出和讨论了综合利用云计算和GPU计算实现高性能、高能效的序列比对平台的研究思路。     

基于CUDA的BLASTN加速研究

目的利用图形处理器（graphicprocessingunit,GPU）计算技术对广泛使用的生物信息学序列比对工具BLASTN加速,服务于新一代测序技术条件下海量生物序列数据分析任务。方法采用计算统一设备架构（corn—puteunifieddevicearchitecture,CUDA）并行计算架构,从GPU多线程并行和多GPU并行两个维度,对核酸序列比对工具BLASTN的种子查找阶段和不允许空位延伸阶段进行并行加速。结果基于CUDA的CUDA—BLASTN取得了显著的加速效果,与FSA．BLAST相比,采用单个NvidiaTeslaC2075显卡在以上两阶段取得了最高达26．8倍的加速比,而且结果准确度没有降低。CUDA—BLASTN特别适合于中长查询序列对长序列数据库的比对任务。结论利用GPU计算可在较大程度上加速序列比对过程,性价比较高,具有很好的应用前景。     

利用Galaxy与高性能计算集群构建本地化一站式生物信息学平台

目的构建本地化的高性能一站式数据分析平台,为生物医学研究的相关科研人员提供便捷高效的计算分析服务。方法将Galaxy软件部署在计算集群上,集成工具软件和数据集;利用分布式资源管理应用接口（DRMAA）实现与SunGridEngine的协同运作,自动调度和分配计算资源;并在集群上构建稳定的weh服务、FrP服务和管理数据库。结果该平台已投入试运行并在不断完善,峰值计算能力达到每秒lO万亿次,存储容量为40TB,提供序列比对、短串映射、基因注释、转录组分析、宏基因组分析及进化分析等多种功能,以及容量约为700GB的人类基因组、病毒、细菌、真菌等参考数据库。结论该平台具备大规模数据分析的能力,能够解决高通量测序所带来的海量生物数据的存储与处理等问题。与在普通服务器上进行数据分析相比,该平台的计算集群能极大地加快数据处理过程,提高研究效率。     

美国航空航天局“人体研究计划”的管理和研究方向

人体研究一直是美国空间生物医学研究的重点方向，“人体研究计划”（HumanResearchProgram，HRP）是美国航空航天局（NASA）启动的重要研究计划，旨在降低影响航天员健康和绩效风险。该文从组织管理、资金投入与研究方向等方面对此项计划进行了简要分析。     

超级刀片计算机上任务并行生物计算程序的效率优化

在后基因组时代，基因结构、功能及其与疾病的关系已成为生物医学的研究热点。在各种实验技术飞速发展的促进下，高复杂度、多元化、海量的生物数据正以指数级的速度不断增长，关于这些数据的存储、计算、分析等已成为严峻挑战。以超级计算机为平台的高性能计算是解决此类难题的有效方法之一。根据对本院超级刀片计算机的使用情况，本文从数据存储、通讯优化、计算效率、资源分配等角度，分析和总结提高任务并行的生物计算程序效率的技术策略，并进一步展望其在生物医学中的应用前景。这些技术策略对于使用同类高性能计算机具有重要的参考价值。     

利用SFU开发Windows和Linux异构环境下的分布式生物信息学应用

Microsoft SFU是一个Windows环境下的高性能UNIX子系统和互操作工具，它允许Windows和UNIX计算机共享数据、安全策略和应用程序。本文通过实例，介绍如何利用SFU将Linux下的Blat软件在Windows系统中重新编译运行，并与Linux计算机共享生物序列数据库，构建异构网络环境下的分布式生物信息学应用。实践证明该技术可集成Windows和Linux的计算能力，提高计算资源的使用效率。     

军事医学科学院生物医学超级计算中心的计算资源与应用

生物信息学是生命科学和信息科学的有机结合，其发展和应用离不开高性能计算资源、技术和方法的支持。本文介绍了我院生物医学超级计算中心的主要计算资源——星盈万亿次超级刀片计算机、支撑软件和生物信息学应用软件，以及我们利用该系统进行大规模生物信息学数据分析的实际计算效果，旨在探讨如何充分挖掘我院生物医学计算资源的潜力，为深入规划我院生物医学研究需要的高性能计算方向提供重要参考。     

计算机仿真技术在生物防护装备研发中的应用

本文介绍了计算机仿真技术在生物防护装备研发过程中的应用。具体讨论了利用基于计算流体力学（computational fluid dynamics,CFD）或晶格玻尔兹曼方法（lattice Boltzmann method,LBM）开发的流体系统仿真软件在对各种生防装备性能进行预测时的要点、难点以及区别。     

生物恐怖事件计算实验支持平台及研究实例

定量分析研究生物恐怖事件过程对国家生物安全战略和有效处置生物恐怖事件具有重要意义。本研究着眼于为定量研究各种类型生物恐怖事件过程提供从基础数据、模型及算法体系-计算实验-情景重现、三维可视化显示的全过程支持,建立生物恐怖事件计算实验支持平台,并以此平台为基础,以北京城区遭受炭疽芽孢杆菌气溶胶袭击为想定进行实例研究并对所造成的危害进行评估。     

外军卫勤研究系列讲座（23） 美陆军医学研究与卫生物资部设置及新研究领域

美陆军医学研究与卫生物资部（USAMRMC）总部位于美国马里兰州的迪特里克堡,是美陆军卫生部（MEDCOM）下属的一个分支机构,也是美陆军军事医学研究的专职管理机构。该机构职能是开展军事医学研究、卫生物资管理、信息技术开发与利用、医疗设施和采办等,为美军官兵提供最佳的预防、治疗和康复等医疗服务。     

Accuracy and efficiency of graphics processing unit (GPU) based Acuros XB dose calculation within the Varian Eclipse treatment planning system

To evaluate, in terms of dosimetric accuracy and calculation efficiency, the implementation of a graphic processing unit (GPU)-based Acuros XB dose calculation engine within version 15.5 of the Varian Eclipse treatment planning system. Initial phantom based calculations and a range of 101 clinical cases were analyzed on a dedicated test system. Dosimetric differences, based on dose-volume histrogram parameters and plan comparison, were compared between central processing unit (CPU) and GPU based calculation. Calculation times were also compared between CPU and GPU, as well as PLAN and FIELD modes. No dosimetric differences were found between CPU and GPU. CPU based calculations ranged from 25 to 533 seconds per plan, reducing to 13 to 70 seconds for GPU. GPU was 4.4 times more efficient than CPU. FIELD mode was up to 1.3 times more efficient than PLAN mode. For the clinical cases and version of Eclipse used, no dosimetric differences were found between CPU and GPU. Based on this, GPU architecture has been safely implemented and is ready for clinical use. GPU based calculation times were superior to CPU, being on average, 4.4 times faster.     

Implementation of GPU accelerated SPECT reconstruction with Monte Carlo-based scatter correction

Objective

Statistical SPECT reconstruction can be very time-consuming especially when compensations for collimator and detector response, attenuation, and scatter are included in the reconstruction. This work proposes an accelerated SPECT reconstruction algorithm based on graphics processing unit (GPU) processing.

Methods

Ordered subset expectation maximization (OSEM) algorithm with CT-based attenuation modelling, depth-dependent Gaussian convolution-based collimator-detector response modelling, and Monte Carlo-based scatter compensation was implemented using OpenCL. The OpenCL implementation was compared against the existing multi-threaded OSEM implementation running on a central processing unit (CPU) in terms of scatter-to-primary ratios, standardized uptake values (SUVs), and processing speed using mathematical phantoms and clinical multi-bed bone SPECT/CT studies.

Results

The difference in scatter-to-primary ratios, visual appearance, and SUVs between GPU and CPU implementations was minor. On the other hand, at its best, the GPU implementation was noticed to be 24 times faster than the multi-threaded CPU version on a normal 128?×?128 matrix size 3 bed bone SPECT/CT data set when compensations for collimator and detector response, attenuation, and scatter were included.

Conclusions

GPU SPECT reconstructions show great promise as an every day clinical reconstruction tool.

     

Evaluation of a new GPU-enabled VMAT multi-criteria optimisation plan generation algorithm

To evaluate the new Varian, graphical processing unit (GPU)-enabled, volumetric-modulated arc therapy (VMAT) multi-criteria optimisation (MCO) tool for both its dosimetric accuracy and calculation time. This is a new capability within V16.0 and greater of the Varian Eclipse treatment planning system that allows VMAT optimisation and dose calculation using the GPU (termed GPU-VMAT). In versions prior to V16.0 VMAT multi-criteria optimisation calculations were only possible using central processing unit (CPU) (termed CPU-VMAT) and Hybrid-VMAT (H-VMAT). The H-VMAT method breaks down the VMAT plan into IMRT fields which utilised GPU calculations. The study consisted of a cohort of 50 patients representing a range of anatomical treatment sites; bladder (5), brain (5), gynae (5), head & neck (5), lung (7), mediastinum (7) prostate (4), oesophagus (7) and rectum (5). Each case was planned to that of a clinical standard (Base) which was compared to a CPU-VMAT, GPU-VMAT and H-VMAT approaches. The study analysed dose to organ at risk (OAR) and target coverage, plan calculation time data and plan complexity through monitor unit (MU) for each approach. Negligible dosimetric differences were found between the CPU-VMAT, GPU-VMAT and H-VMAT approaches for the cohort of patients evaluated. The largest dosimetric change were observed in the lacrimal gland for a head and neck case, where the GPU-VMAT and H-VMAT achieved a max dose of +2.8 ± 0.0 Gy and −4.6 ± 0.0 Gy, respectively, when compared to CPU-VMAT. The majority of organ at risk’s (OAR) provided indistinguishable dosimetric outcomes, namely: heart, kidneys, femur, lens, oral cavity and oesophagus. Large time savings were found using the GPU-VMAT technique compared to CPU-VMAT, a mean decrease in calculation time across all sites of 60.2% ± 15.6%. Negligible dosimetric change between the 2 techniques and large time saving were observed with the GPU-VMAT and H-VMAT approaches when compared to the CPU-VMAT. We have shown that the GPU-VMAT technique has been safely implemented with minimal differences from CPU-VMAT, but with significant optimisation and calculation times savings.     

Evaluation of a new hybrid VMAT-IMRT multi-criteria optimization plan generation algorithm

To evaluate the Varian ‘Fast hybrid multi-criteria optimization (MCO) volumetric modulated arc therapy (VMAT)’ (H-VMAT) tool for both its dosimetric accuracy and calculation time. This is a new function within V15.6 of the Varian Eclipse treatment planning system that allows VMAT optimization and dose calculation using the graphical processing unit (GPU). In versions prior to V15.6 VMAT MCO calculations were only possible using central processing unit (CPU) not GPU. We termed this approach as native VMAT (N-VMAT). The study consisted of a cohort of 53 patients representing a range of anatomical treatment sites; bladder (5), brain (6), gynaecological (5), head & neck (5), lung (7), mediastinum (7) prostate (6), oesophagus (7), and rectum (5). Each case was planned to that of a clinical standard (Base) which was compared to a H-VMAT and N-VMAT approach. The study analyzed plan calculation time data, dose to organ at risk (OAR) and target coverage for each approach. Negligible dosimetric differences were found between the H-VMAT and N-VMAT approach for the cohort of patients evaluated. The largest dosimetric changes where observed in the optic chiasm and lacrimal gland where the H-VMAT achieved a max dose of 50.9 ± 7.7 Gy and 8.0 ± 0.5 Gy in comparison to the N-VMAT 53.1 ± 6.3 Gy and 10.2 ± 2.9 Gy, respectively. Several OAR's provided indistinguishable dose outcomes, namely; brainstem, heart, kidney's, lens, parotid, and spinal cord. Large time savings were found using the H-VMAT technique when compared to N-VMAT, being 5 to 40 times faster or up to 75 minutes time saving (average of 25 minutes). Negligible dosimetric change between the 2 techniques and large time savings were observed with the GPU enabled H-VMAT approach. We have shown that the H-VMAT technique has been safely implemented and is ready for clinical use.     

High resolution in vivo micro-CT with flat panel detector based on amorphous silicon

A high-resolution in vivo micro-CT system for combining with fluorescence molecular tomography (FMT), was constructed and applied in small animal imaging. The fast scanning micro-CT system designed to provide high-resolution anatomic information and reconstruction priors, consisted of a flat panel detector (FPD) based on amorphous silicon (a-Si) and a micro-focus x-ray tube. The Feldkamp algorithm was adopted in image reconstruction with graphic processing unit (GPU). The system spatial resolution of 13 lp/mm was achieved when the diameter of image field was 6 cm with the system magnification factor of 4. No obvious beam hardening artifact was observed in transaxial image of a water phantom after correction. The contrast-to-noise ratio (CNR) study of various tissue phantoms was also presented. The in-vivo imaging of an anesthetic mouse was performed to demonstrate the feasibility of our system.     

Image processing on Macintosh II: a practical boundary finding algorithm for biomedical measurement

With increasing research interest in displaying and analyzing biomedical images, a practical personal computer based off-line image processing software would be useful. This paper describes the implementation of an image processing work station on a Macintosh II which features a novel edge detection capability useful for biomedical measurement. The boundary finding algorithm is coded in Turbo Pascal, and operates at a speed comfortable for interactive operation. Depending on the complexity of the problem, it usually takes less than a minute for the measurement of an image. The edge detection algorithm has an built-in edge detector with decision-making capability, and can be efficiently controlled by a mouse. In this way, the local accuracy of an automatic edge detection operator, and the global accuracy of the human eye (through manual control of a mouse) are combined.     

Biomedical research: its mission within the French Military Medical Service.

The French Military Medical Service is organized as a distinct corps to support Army, Navy, and Air Force operations. This complex mission is accomplished through five operational components: (1) direct medical support of the force units; (2) hospital nursing and expertise; (3) biomedical research; (4) biomedical training; and (5) medical supply. Additionally, the French Military Medical Service is committed to humanitarian and civil medical support. Advanced biomedical research, particularly on infectious diseases and treatment of injuries, is actively pursued. Fundamental and applied research is needed to anticipate potential threats and improve medical support and care of French forces. The importance of biomedical research was recognized as necessary to develop technological improvements in rescue operations and to provide the military command with scientifically based advice. Biomedical military research has often been the engine of progress in medicine and surgery. Chief among those developments has been a special emphasis on infectious diseases and wound treatment.