[名稱]: 實作網路版引子探針設計演算軟體的圖形處理器演算統合裝置架構核心世衛組織(WHO)於1995 年建議醫師針對感染病患症狀看診紀錄病歷，再由臨床檢驗室針對病歷症狀確認病因的感染症狀相關病原；問題在於類症病原群組的鑑別檢測，依據傳統平台逐項實施窒礙難行，建立整群同步鑑別的創新檢測平台自然成為亟需且必要的應用研究目標，藉以針對已知類症病原群組提供鑑別檢測實務，整體快速區分重症與輕症病毒的重要風險資訊。本項研究計畫研擬開發達成目標的軟體組合解決方案，並充分符合國科會資訊工程(二)學門重點規劃主題(1)服務導向計算與應用：1.科技化服務與應用a. 生命照護：健康管理、醫療照護、生醫資訊應用。針對建立網路版軟體組合議題，依據自有的個人電腦版引子探針設計演算[PDA]與修飾版簡要型遺傳演算法[MCGA]核心技術，整合(i)軟體演算統合裝置架構[CUDA]核心與(ii)硬體圖形處理器[GPU]的解決方案，克服個人電腦版演算速度耗時不適提供網路開放應用的難題。本項研究計畫研擬實作的網路版PDA-CUDA 平行化修飾版簡要型遺傳演算法[pMCGA]軟體平台，擴展自有的個人電腦版核心技術的引子探針設計演算[PDA]與修飾版簡要型遺傳演算法[MCGA]。本項應用研究計劃的PDA-CUDA 技術平台成果產出，合計研擬申請一案美國發明專利，與投稿一篇SCI 期刊文獻。 [Title]: Implement Web-version Primer/Probe Design Algorithm (PDA) with Compute Unified Device Architecture (CUDA) Kernels on Graphic Processing Unit (GPU) WHO officially required physicians to record manifested clinical syndromes of patients on infectious diseases while to subsequently confirm the syndrome-related pathogens with clinical laboratory. The obvious differential detection as indicated may demand innovative platform on whole-set basis for same-time detection since that conventional itemized detection demonstrated only failure in feasibility. The differential detection platform became an application goal in order to provide important clinical information of practical risks such as severe-disease or mild-disease virus. This research project intends to develop the software package of integrated solutions which complied with NSC’s discipline code of EB(15) or [E0819] Bioinformatics on priority theme [(1).1.a] Biomedical information application pertaining to (a) life-care’s health management, medical care, and biomedical informatics application. Towards the issue of implementing web-version software package, feasible goal-tending tasks maybe based on our in-house PC-version algorithms of primer/probe-design algorithm [PDA] and modified compact genetic algorithm [MCGA]. That is, the integration of both software of CUDA kernel and hardware of GPU may overcome the barrier for PDA web service due to severe consumption of computation time. The proposed software implementation for web-version PDA-CUDA of parallel MCGA [pMCGA] is the extension from our lab-owned algorithms of primer/probe-design algorithm [PDA] and modified compact genetic algorithm [MCGA]. On evaluating performance, this proposed PDA-CUDA project of applied research shall file-in one USPTO patent applications of invention and as well shall submit one publication draft to SCI journal.
