벼 유전자원의 수당립수 증진 유전자 유전형 분석

신동진,김태헌,이지윤,조준현,이종희,송유천,박동수,오명규,Shin, Dongjin,Kim, Tae-Heon,Lee, Ji-Yoon,Cho, Jun-Hyun,Lee, Jong-Hee,Song, You-Chun,Park, Dong-soo,Oh, Myeong-Kyu 한국식물생명공학회 2017 식물생명공학회지 Vol.44 No.4

벼 수량성 증진을 위하여 수당립수 증진 유전자로 보고된 5종의 유전자에 대한 분자표지를 검정하고 유전자원 479점에서 이들의 유전자에 대한 유전형을 검정하였다. 판독이 용이한 Gn1a및 DEP1, Apo1유전자의 In/del 분자표지를 각각 개발하였고 Ghd7과 Nal1 유전자에 대하여서는 기존 보고된 SNP 분자표지를 이용하여 편리성을 검정하였다. 이들 분자표지는 아가로즈젤에서 각각의 유전형 판독이 용이하기에 벼 수량성 향상을 위한 분자육종에 적용이 가능할 것으로 기대되었다. 유전자원 479점에서 수당립수 증진 유전자 5종의 유전형을 분석하였을 때 총 13개의 haplogype으로 분류되었다. 대부분의 Indica 품종과 Japonica 품종은 haploptype 1과 haplotype 13에 속하였다. 나머지 haplotype에 속한 55점의 유전자원은 수당립수 증진 유전자에 대한 유전다양성을 보유한 자원으로 유전체 분석 등을 위한 핵심집단으로 활용할 수 있을 것으로 판단되었다. 유전자원 396점의 수량구성요소를 비교하였을 때, Nal1을 제외한 4종의 수당립수 증진 유전자의 수량증진 대립유전자형에서 이삭 수가 0.6 ~ 0.8개/주 감소하였으나 수당립수는 이삭당 27 ~ 29개 증진되었다. Nal1 유전자는 유전적 배경에 따라 효과가 다르게 나타나며, Nal1-japonica 대립유전자형의 수당립수 증진 효과보다 Nal1-indica 대립유전자형이 감소효과가 큰 것으로 추측되었다. 앞으로 본 논문에서 검정된 수당립수 증진 분자표지 5종과 유전자원의 유전형을 정보를 바탕으로 벼 수량성 증진 육종에 활용하고자 한다. ARice is an important staple food in the world and rice yield is one of the main traits for rice breeding. Several genes involved in increasing the yield have been identified through map-based gene cloning within natural variations in rice. These identified genes are good targets for introducing a genetic trait in molecular breeding. Here, we chose five genes reported to be involved in increasing grain number per panicle in rice; Gn1a, dep1, Apo1, Ghd7, and Nal1. We developed In/Del markers for Gn1a, and dep1, Apo1, and applied the reported SNP markers for Ghd7 and Nal1. We were easily able to examine the genotype of each gene on agarose gel. We tested the genotypes on 479 rice resources that we held with evaluated molecular markers. According to the genotype of each gene, rice resources were divided into 13 haplotypes, and most of the Indica and Japonica varieties were included in haplotypes 1 and 13, respectively. When we examined the effect of each gene on grain number per panicle and panicle number per plant, panicle number per plant in the yield negative allele group for each gene was reduced by approximately 0.3 to 0.8 compared to that in the yield positive allele group. However, the number of yield positive alleles for each gene was higher by about 21 to 27 grains per panicle than that of yield negative alleles. Although most of the varieties were grouped in haplotypes 1 and 13, we believe that this genotype information with evaluated molecular markers will be useful in rice breeding for increasing the yield with grain number per panicle.

실도로 연비 향상을 위한 경제운전 안내시스템 개발

신동진(Dongjin Shin),윤헌중(Heonjoong Yoon),배현철(Hyuncheol Bae),송경근(Kyungkeun Song) 한국자동차공학회 2009 한국자동차공학회 학술대회 및 전시회 Vol.2009 No.11

Eco Drive indicating system lets driver know whether car is operating in fuel-efficient state or not, so that the driver is guided to save fuel. Eco Drive state is distinguished based on A/T shift pattern data, engine operating status, transmission operating status. With Eco Drive indicating system, it was proved that fuel-efficiency can be increased by 4~26%.

웹 서비스에 환경에서 빅데이터 기반 이상행위 탐지 기술

신동진(Dongjin Shin),이호근(Hogeun Lee),은종민(Jongmin Eun),김정준(Jeongjoon Kim),박정민(JeongMin Park) 인문사회과학기술융합학회 2018 예술인문사회융합멀티미디어논문지 Vol.8 No.4

본 논문은 웹 서비스의 로그와 게시판 데이터베이스 정보를 수집하고 이상행위에 대한 분석을 위한 빅데이터 기반의 웹 서비스 로그와 게시판 분석 연구에 관해 소개한다. 기존의 웹 서비스의 필터링 방식은 관리자가 직접 수 많은 데이터들을 모니터링 후 특정 키워드를 등록하고, 악성 사용자를 직접 등록해서 필터링하는 방식으로 진행되었다. 하지만 이 방식은 사람이 직접 키워드 업데이트를 해주어야 하고, 스팸 키워드에 일치해야만 필터링이 되어 조금만 다른 키워드로 글을 등록하면 필터링이 되지 않았다. 따라서 본 논문에서는 기존 관리자가 게시판을 직접 모니터링하면서 필터링 작업을 하는 방식보단 빅데이터를 활용한다면 다량의 데이터를 실시간으로 처리해 속도를 향상 시킬 수 있으며 분석에 대해 정확도가 높은 장점이 있다. 빅데이터는 오늘날 다양한 분야에 활용되는 데이터수집 및 분석기술의 경향으로, 정형 데이터와 비정형 데이터 분석을 요구하는 다양한 분야에 접목되어 효용성을 인정받고 있다. 웹 서비스의 로그와 게시판 역시 정형과 비정형 데이터 형식으로 로그 데이터는 Flume을 통해 실시간 저장되며 Sqoop은 필요한만큼 대량의 데이터 하둡에 저장한다. 저장된 데이터들은 빅데이터의 처리 도구인 Hive와 Pig를 통하여 최종적으로 R프로그래밍을 통해 분석이 완료된다. 이처럼 빅데이터와 접목 시 좀 더 원활하고 정확한 필터링이 가능하며 이상행위를 탐지하여 빅데이터 기반으로 웹 서비스 환경을 분석한다. This paper collect Web Services log, board database information and introduces a Big Data-based web service log, board and analysis for analysis of abnormal behaviors. The methods of filtering existing web services were conducted by the administrator to register a large number of data, register specific keywords, register, filter malicious users manually. However, this method has to be filtered directly by the person in question, and it is not filtered by filtering that it is filtered only by the word spam and is filtered to a slightly different keyword. Thus, in this thesis, if the existing administrator uses big data instead of monitoring the bulletin boards, it can be processed in real time by utilizing large amounts of data in real time to speed up and improve the accuracy of the analysis. Big data is widely used in diverse fields of data gathering and analytics today, garnering benefits in diverse fields requiring structured data and unstructured data analysis. Log and bulletin boards of web services are also stored in structured and unstructured formats in structured and unstructured formats, and Sqoop is stored in large amounts of data Hadoop as needed. Stored data is finally analyzed through R programming through Hive and Pig, the processing tool of Big Data. Big data, more accurate filtering, and more accurate filtering of big data, which detects the behavior of a web service and analyzes the web service environment based on big data.

애플스토어 디자인에 나타난 신고전주의 건축의 형태 및 공간 특성 - 애플스토어 1기 사례를 중심으로 -

신동진 ( Shin Dongjin ),김종진 ( Kim Jongjin ) 한국공간디자인학회 2021 한국공간디자인학회논문집 Vol.16 No.8

(Background and Purpose) Apple Store has been demonstrating its innovative design over the past 20 years, and is a typical case to prove that even its retail store can be evaluated as a world-class building. The purpose of this study is to suggest that the archetype of “simplicity” demonstrated in the Apple Store design is derived from “Neo-Classicism architecture style.” Accordingly, the characteristics of two types in the architecture of Neo-Classicism were reviewed, and how its architectural form and space characteristics were applied to Apple Store design were analyzed. (Method) In this study, the research method was categorized into literature reviews and case studies. The literature reviews attempted to determine the association between the design of Apple Store and the architecture of Neo-Classicism on the basis that Bauhaus architects mentioned in Steve Jobs’ biography were influenced by the ideas and form of Neo-Classicism architecture. Correspondingly, two types of Neo-Classicism architecture, revivalistic Neo-Classicism architecture and revolutional Neo-Classicism architecture, and related Modern Rationalism architects were examined through previous studies. Furthermore, by synthesizing the characteristics derived as the result and the characteristics of two types, an aboveground store located on a street and an underground store located in a square, in the design of Apple Store, a framework for analysis to be used in subsequent case studies was presented. For the case study, six remarkable stores, three for each type, were selected among flagship concept stores that entirely reflected the design philosophy of Jobs before the design strategy of Apple Store was changed in 2017. For aboveground stores located on the street, Apple Stanford Shopping Center, Apple Omotesando, and Apple Westlake were considered, and for underground stores located in square, Apple Fifth Avenue in New York, Apple Pudong, and Apple Zorlu Center were chosen. (Results) Through an analysis based on the framework of the case analysis presented above, the study attempted to illuminate Apple store design from the architectural history perspective by clarifying that the characteristics of two types in Neo-Classicism architecture were transferred to two types of Apple Store design. (Conclusions) This study tried to determine the architectural historical archetype shown in the first period cases of Apple Store design representing Jobs' ideological world into a real world. Apple Store strongly symbolizes and imprints its brand and product through its architecture. In addition, the interior space of Apple Store offers consumers a special experience to immerse themselves in its brand and products. It is interesting that Apple Store, a symbol of innovation in the 21st century, resembles the classical architecture of the 18th century. This is meaningful as a new perspective to consider the architecture and space of retail stores at a higher level.

입력 영상과 후처리를 이용한 변이 지도 영상의 개선 및 객체 분할

신동진(Dongjin Shin),최민수(Minsoo Choi),한동일(Dongil Han) 대한전자공학회 2006 대한전자공학회 학술대회 Vol.2006 No.11

Robots have been used to perform in many ways for a long time. To be used to perform like these various tasks, they are supposed to be equipped with stereo vision system which is able to segment objects and to calculate distance information. These tasks could be performed by calculating the distance and object segmentation through the stereo vision system. But, there are lots of noises and errors in real life of disparity map image which is captured by using a stereo camera so we can’t trust the result obtained from this disparity map image. In this paper, we present a method which is able to improve the stereo image by comparing with input image. Then we propose a method which segments the object by using the improved disparity map images.

신세포암 분화도 판정의 정확도 향상을 위한 Handcrafted 특징 선택 기법

신동진(Dongjin Shin),조현준(HyunJun Cho),정승원(Seung-Won Jung) 대한전자공학회 2024 대한전자공학회 학술대회 Vol.2024 No.6

In this paper, we propose a novel method for extracting and analyzing nuclear image features to improve the accuracy and reproducibility of diagnoses using the Renal Cell Carcinoma (RCC) Fuhrman grading system. We extracted morpho-logical characteristics of nuclei and selected optimal features through correlation analysis and the Minimum redundancy maximum relevance technique. The validity of these features was confirmed through ANOVA and Kendall’s rank correlation analyses. Our experimental results, using the KMC public dataset, demonstrate that this approach reduces the subjectivity of the Fuhrman grading system. By providing objective criteria for pathological diagnosis, our selected features can enhance the accuracy of RCC diagnoses.

의무지원체계의 단일화시 수의병과의 역할

신동진 ( Dongjin Shin ),금현옥 ( Hyunok Keum ),박경완 ( Kyoungwan Park ) 국군의무사령부 2008 대한군진의학학술지 Vol.39 No.1

Survey of Distributed Training in Imperative Deep Learning Frameworks

DongJin Shin(신동진),Byung-Gon Chun(전병곤) 한국정보과학회 2018 한국정보과학회 학술발표논문집 Vol.2018 No.12

제어모델 코드생성 기반 로직 프로토타이핑 환경 개발

신동진(Dongjin Shin),정명규(Myunggyu Jung),이동열(Dongyeol Lee),서재암(Jaeam Seo) 한국자동차공학회 2018 한국자동차공학회 부문종합 학술대회 Vol.2018 No.6

Logic prototyping environment based on model code generation is developed in VDMS(Vehicle Data Monitoring System). C code generation using Matlab Embedded Coder, matching S/W between model input/output and CAN signal, realtime task scheduler environment running on VDMS application processor is implemented. This paper examines the process and result of the VDMS logic prototyping environment.

Works - TAILWIND

이선민,신동진,Lee Sunmin,Shin Dongjin Korea Institute of Registered Architects 2023 建築士 Vol.653 No.-