http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Minimal genome: Worthwhile or worthless efforts toward being smaller?
Choe, Donghui,Cho, Suhyung,Kim, Sun Chang,Cho, Byung‐,Kwan WILEY‐VCH Verlag 2016 BIOTECHNOLOGY JOURNAL Vol.11 No.2
<P><B>Abstract</B></P><P>Microbial cells are versatile hosts for the production of value‐added products due to the well‐established background knowledge, various genetic tools, and ease of manipulation. Despite those advantages, efficiency of newly incorporated synthetic pathways in microbial cells is frequently limited by innate metabolism, product toxicity, and growth‐mediated genetic instability. To overcome those obstacles, a minimal genome harboring only the essential set of genes was proposed, which is a fascinating concept with potential for use as a platform strain. Here, we review the currently available artificial reduced genomes and discuss the prospects for extending use of the genome‐reduced strains as programmable chasses. The genome‐reduced strains generally showed comparable growth to and higher productivity than their ancestral strains. In <I>Escherichia coli</I>, about 300 genes are estimated as the minimal number of genes under laboratory conditions. However, recent advances revealed that there are non‐essential components in essential genes, suggesting that the design principle of minimal genomes should be reconstructed. Current technology is not efficient enough to reduce large amount of interspaced genomic regions or to synthesize the genome. Furthermore, construction of minimal genome frequently has failed due to lack of genomic information. Technological breakthroughs and intense systematic studies on genomes remain tasks.</P>
STATR: A simple analysis pipeline of Ribo-Seq in bacteria
Donghui Choe,Bernhard Palsson,Byung-Kwan Cho 한국미생물학회 2020 The journal of microbiology Vol.58 No.3
Gene expression changes in response to diverse environmental stimuli to regulate numerous cellular functions. Genes are expressed into their functional products with the help of messenger RNA (mRNA). Thus, measuring levels of mRNA in cells is important to understand cellular functions. With advances in next-generation sequencing (NGS), the abundance of cellular mRNA has been elucidated via transcriptome sequencing. However, several studies have found a discrepancy between mRNA abundance and protein levels induced by translational regulation, including different rates of ribosome entry and translational pausing. As such, the levels of mRNA are not necessarily a direct representation of the protein levels found in a cell. To determine a more precise way to measure protein expression in cells, the analysis of the levels of mRNA associated with ribosomes is being adopted. With an aid of NGS techniques, a single nucleotide resolution footprint of the ribosome was determined using a method known as Ribo- Seq or ribosome profiling. This method allows for the highthroughput measurement of translation in vivo, which was further analyzed to determine the protein synthesis rate, translational pausing, and cellular responses toward a variety of environmental changes. Here, we describe a simple analysis pipeline for Ribo-Seq in bacteria, so-called simple translatome analysis tool for Ribo-Seq (STATR). STATR can be used to carry out the primary processing of Ribo-Seq data, subsequently allowing for multiple levels of translatome study, from experimental validation to in-depth analyses. A command- by-command explanation is provided here to allow a broad spectrum of biologists to easily reproduce the analysis.
최적화된 쿼드트리를 이용한 2차원 연기 데이터의 효율적인 슈퍼 해상도 기법
최유연(YooYeon Choe),김동희(Donghui Kim),김종현(Jong-Hyun Kim) 한국컴퓨터정보학회 2021 한국컴퓨터정보학회 학술발표논문집 Vol.29 No.1
본 논문에서는 SR(Super-Resolution)을 계산하는데 필요한 데이터를 효율적으로 분류하고 분할하여 빠르게 SR연산을 가능하게 하는 쿼드트리 기반 최적화 기법을 제안한다. 제안하는 방법은 입력 데이터로 사용하는 연기 데이터를 다운스케일링(Downscaling)하여 쿼드트리 연산 소요 시간을 감소시키며, 이때 연기의 밀도를 이진화함으로써, 다운스케일링 과정에서 밀도가 손실되는 문제를 피한다. 학습에 사용된 데이터는 COCO 2017 Dataset이며, 인공신경망은 VGG19 기반 네트워크를 사용한다. 컨볼루션 계층을 거칠 때 데이터의 손실을 막기 위해 잔차(Residual)방식과 유사하게 이전 계층의 출력 값을 더해주며 학습한다. 결과적으로 제안하는 방법은 이전 결과 기법에 비해 약15~18배 정도의 속도향상을 얻었다.