• International Journal of Contents
• /
• 제8권1호
• /
• pp.23-29
• /
• 2012

Multiple expression levels of genes obtained using time series microarray experiments have been exploited effectively to enhance understanding of a wide range of biological phenomena. However, the unique nature of microarray data is usually in the form of large matrices of expression genes with high dimensions. Among the huge number of genes presented in microarrays, only a small number of genes are expected to be effective for performing a certain task. Hence, discounting the majority of unaffected genes is the crucial goal of gene selection to improve accuracy for disease diagnosis. In this paper, a non-Gaussian weight matrix obtained from an incremental model is proposed to extract useful features of multivariate time series microarrays. The proposed method can automatically identify a small number of significant features via discovering hidden variables from a huge number of features. An unsupervised hierarchical clustering representative is then taken to evaluate the effectiveness of the proposed methodology. The proposed method achieves promising results based on predictive accuracy of clustering compared to existing methods of analysis. Furthermore, the proposed method offers a robust approach with low memory and computation costs.

• Genomics & Informatics
• /
• 제5권3호
• /
• pp.113-117
• /
• 2007

The development of microarray technology has facilitated the understanding of gene expression profiles. Despite its convenience, the cause of dye-bias that confounds data interpretation in dual-color DNA microarray experiments is not well known. In order to economize time and money, it is necessary to identify the cause of dye bias, since designing dye-swaps to reduce the dye-specific bias tends to be very expensive. Hence, we sought to determine the reliable cause of systematic dye bias after treating murine macrophage RAW 264.7 cells with 2-keto-3-deoxyoctonate (KDO), interferon-beta $(IFN-{\beta})$, and 8-bromoadenosine (8-BR). To find the cause of systematic dye bias from the point of view of fluorescence quenching, we examined the correlation between systematic dye bias and the proportion of each nucleotide in mRNA and oligonucleotide probe sequence. Cy3-dye bias was highly correlated with the proportion of adenines. Our results support the fact that systematic dye bias is affected by fluorescence quenching of each feature. In addition, we also found that the strength of fluorescence quenching is based on not only dye-dye interactions but also dye-nucleotide interactions as well.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2006년도 Principles and Practice of Microarray for Biomedical Researchers
• /
• pp.78-78
• /
• 2006

Gene expression analysis can be performed by one-color (intensity-based) or two-color (ratio-based) microarray platforms depending on the specific applications and needs of the researcher. The traditional two-color approach is well founded from a historical and scientific standpoint, and the one-color approach, when paired with high quality microarrays and a robust workflow, offers additional flexibility in experimental design. Two of the major requirements of any microarray platform are system reproducibility, which provides the means for high confidence experiments and accurate comparison across multiple samples; and high sensitivity, for the detection of significant gene expression changes, including small fold changes across multiple gene sets. Each of these requirements is fulfilled by the Agilent One-color Gene Expression Platform as illustrated by the data included in this study. As a result, researchers have the ability to take advantage of the enhanced performance and sensitivity of Agilent's 60-mer oligonucleotide microarrays, and experience the first commercial microarray platform compatible with both one- and two-color detection.

• 한국잠사학회:학술대회논문집
• /
• 한국잠사학회 2003년도 제46회 춘계 학술연구 발표회
• /
• pp.47-48
• /
• 2003

The development of cDNA microarray has permitted the analysis of thousands of genes simultaneously. cDNA microarray has been used to analyze gene expression profiles during developmental stage in both single and multicellular organisms. Two significant factors contributing to the limitation of the development of cDNA microarray in the Bombyx mori are the shortage of accessible repositories of cDNA clones and ESTs and the relative scarcity of facilities to produce microarrays and analyze the data generated. (omitted)

• 대한약침학회지
• /
• 제11권4호
• /
• pp.5-14
• /
• 2008

Objective This study was undertaken to determine the effect of Whallak-tang on expression of CD/cytokine Genes. Methods The expression of CD/Cytokine Genes were examined by cDNA microarray using the human mast cell line(HMC-1). Results The expression of ATP5F1, FLJ20671, unknown, KIAA0342, OAS2, unknown genes were increased in $200{\sim}300%$ range. The expression of unknown, MDS006, IFITM1, MRPL3, ZNF207, FTH1, FBP1, NRGN, NR1H2, KIAA0747 genes were decreased in $0{\sim}33%$ range. Conclusion These results would provide important basic data on the possibility of the clinical treatment of Whallak-tang in musculoskeletal disease.

• ETRI Journal
• /
• 제41권3호
• /
• pp.358-370
• /
• 2019

The main aim of this study is to select the optimal set of genes from microarray cancer datasets that contribute to the prediction of specific cancer types. This study proposes the enhancement of the feature selection filter algorithm based on Joe's normalized mutual information and its use for gene selection. The proposed algorithm is implemented and evaluated on seven benchmark microarray cancer datasets, namely, central nervous system, leukemia (binary), leukemia (3 class), leukemia (4 class), lymphoma, mixed lineage leukemia, and small round blue cell tumor, using five well-known classifiers, including the naive Bayes, radial basis function network, instance-based classifier, decision-based table, and decision tree. An average increase in the prediction accuracy of 5.1% is observed on all seven datasets averaged over all five classifiers. The average reduction in training time is 2.86 seconds. The performance of the proposed method is also compared with those of three other popular mutual information-based feature selection filters, namely, information gain, gain ratio, and symmetric uncertainty. The results are impressive when all five classifiers are used on all the datasets.

• 한국어병학회지
• /
• 제36권2호
• /
• pp.191-211
• /
• 2023

Rock bream iridovirus (RBIV) causes high mortality and economic losses in rock bream (Oplegnathus fasciatus) aquaculture industry in Korea. Although, the immune responses of rock bream under RBIV infection have been studied, there is not much information at the different stages of infection (initial, middle and recovery). Gene expression profiling of rock bream under different RBIV infection stages was investigated using a microarray approaches. In total, 5699 and 6557 genes were significantly up- or down-regulated over 2-fold, respectively, upon RBIV infection. These genes were grouped into categories such as innate immune responses, adaptive immune responses, complements, lectin, antibacterial molecule, stress responses, DNA/RNA binding, energy metabolism, transport and cell cycle. Interestingly, hemoglobins (α and β) appears to be important during pathogenesis; it is highly up-regulated at the initial stage and is gradually decreased when the pathogen most likely multiplying and fish begin to die at the middle or later stage. Expression levels were re-elevated at the recovery stage of infection. Among up-regulated genes, interferon-related genes were found to be responsive in most stages of RBIV infection. Moreover, X-linked inhibitor of apoptosis (XIAP)-associated factor 1 (XAF1) expression was high, whereas expression of apoptosis-relate genes were low. In addition, stress responses were highly induced in the virus infection. The cDNA microarray data were validated using quantative real-time PCR. Our results provide novel inslights into the broad immune responses triggered by RBIV at different infection stages.

• Reproductive and Developmental Biology
• /
• 제32권4호
• /
• pp.229-235
• /
• 2008

Silencing of Dicer1 by siRNA did not inhibit development up to the blastocyst stage, but decreased expression of selected transcription factors, including Oct-4, Sox2 and Nanog, suggesting that Dicer1 gene expression is associated with differentiation processes at the blastocyst stage (Cui et al., 2007). In order to get insights into genes which may be linked with microRNA system, we compared gene expression profiles in Gapdh and Dicer1 siRNA-microinjected blastocysts using the Applied Biosystem microarray technology. Our data showed that 397 and 737 out of 16354 genes were up- and down-regulated, respectively, following siRNA microinjection (p<0.05), including 24 up- and 28 down-regulated transcription factors. Identification of genes that are preferentially expressed at particular Dicer1 knock down embryos provides insights into the complex gene regulatory networks that drive differentiation processes in embryos at blastocyst stage.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2006년도 Principles and Practice of Microarray for Biomedical Researchers
• /
• pp.20-28
• /
• 2006

최근의 microarray 기술의 발달로 인해 점점 더 많은 양의 mRNA 발현 데이터가 쌓여 가고 있다. 이제는 데이터를 만드는 단계보다는 데이터로부터 중요한 생물학적 의미를 끌어내는 것이 더욱 중요한 일이 되었다. micorarray 기술이 처음 도입된 이후로, 많은 앨고리즘과 소프트웨어가 개발되어, 실험자들이 microarray 데이터로부터 생물학적 의미를 끌어내는 작업을 도와주어 왔다. 그런데, 이전의 데이터 마이닝 방법들은 거의 예외 없이 전체 데이터로부터 선택된 몇 십, 몇 백 개의 유전자 리스트로부터 출발한다. 그런데, 이러한 방법 (over-representation analysis, ORA로 줄임)은 몇 가지 한계를 가지고 있어서, 최근에는 전체 데이터로부터 의미 있는 유전자 세트 (gene set)를 찾아내는 방법들이 도입되었다. 본 세미나는 이런 방법들, 줄여서 gene set analysis라 함, 에 사용되는 앨고리즘들과 소프트웨어들을 비교, 검토하고자 한다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2003년도 가을 학술발표논문집 Vol.30 No.2 (2)
• /
• pp.733-735
• /
• 2003

DNA Microarray 기술은 유전자의 발현여부를 매우 빠르게 검사할 수 있는 도구이며 각종 질병의 발생여부를 예측하기 위한 정보를 제공한다. 유전자 발현 데이터로부터 암의 발생 여부를 예측하기 위해서는 기존의 접근방법과 다른 기계학습 기법이 요구된다. 일반적으로 샘플의 개수가 극히 적은 반면에 특징의 개수는 수천에서 수만 개가 존재하기 때문에 문제의 특성에 맞는 분류기의 구조를 결정하는 것이 매우 어려운 일이기 때문이다. 진화 신경망은 신경망의 구조와 가중치를 동시에 학습하며 사용자는 각 개체의 적합도를 평가할 수 있는 방법만 제공해 주면된다. 특히 신경망의 구조를 사전에 고정하지 않아도 되는 장점이 있기 때문에 전문적인 지식이 없는 사용자라도 이용가능하다. 대장암 데이터에 대한 실험결과 제안하는 분류기 모델이 다층 퍼셉트론, SVM (support vector machine), 최근접 이웃 방법에 비해 향상된 성능을 보였다.