東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Probe-level microarray analyses: Ha...

Fang, Xuemin.

FindBook

Google Book

Amazon

博客來

Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes./
作者:	Fang, Xuemin.
面頁冊數:	107 p.
附註:	Source: Dissertation Abstracts International, Volume: 65-10, Section: B, page: 4933.
Contained By:	Dissertation Abstracts International65-10B.
標題:	Biology, Biostatistics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3149541
ISBN:	0496094017

Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes.
Fang, Xuemin.

Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes. - 107 p.

Source: Dissertation Abstracts International, Volume: 65-10, Section: B, page: 4933.

Thesis (Ph.D.)--Harvard University, 2004.

A variety of system errors could arise from large-scale gene expression microarrays, among them one of the most severe problems is saturation. Saturation takes place when there is an abundant amount of transcripts during the hybridization, resulting in probe pixel fluorescence that exceeds the response range of the laser scanner. The intensity measurement is then truncated to a lower value, usually the value of the maximum intensity scale of the scanner. Saturation has several severe effects on the estimated expression indexes. In the simplest case, when both PM and MM probes are saturated, their difference is zero, and the gene expression index may be estimated to be zero with many current statistical methods. We impute the saturated probe intensities with their conditional expectation based on a normal distribution assumption, and this helps to reduce the saturation effect to a large extent.

ISBN: 0496094017Subjects--Topical Terms:

1018416
Biology, Biostatistics.

Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes.
LDR:03616nmm 2200277 4500 001 1839615
005 20050624104422.5
008 130614s2004 eng d
020 $a 0496094017
035 $a (UnM)AAI3149541
035 $a AAI3149541
040 $a UnM $c UnM
100 1 $a Fang, Xuemin. $3 1927994
245 1 0 $a Probe-level microarray analyses: Handling of probe intensity saturation; a random effect model to estimate cross-hybridization in expression index computation; using probe sequence to identify contaminating genes.
300 $a 107 p.
500 $a Source: Dissertation Abstracts International, Volume: 65-10, Section: B, page: 4933.
500 $a Adviser: Wing Hung Wong.
502 $a Thesis (Ph.D.)--Harvard University, 2004.
520 $a A variety of system errors could arise from large-scale gene expression microarrays, among them one of the most severe problems is saturation. Saturation takes place when there is an abundant amount of transcripts during the hybridization, resulting in probe pixel fluorescence that exceeds the response range of the laser scanner. The intensity measurement is then truncated to a lower value, usually the value of the maximum intensity scale of the scanner. Saturation has several severe effects on the estimated expression indexes. In the simplest case, when both PM and MM probes are saturated, their difference is zero, and the gene expression index may be estimated to be zero with many current statistical methods. We impute the saturated probe intensities with their conditional expectation based on a normal distribution assumption, and this helps to reduce the saturation effect to a large extent.
520 $a In Affymetrix GeneChip experiments it is not uncommon to encounter situations where a designed probe not only hybridizes to its target transcript (gene), but also hybridizes to other non-target gene(s). This activity is called cross-hybridization. Cross-hybridization makes it difficult to measure the target gene expression indexes from the probe intensities. In this paper we proposed a random effect model approach to incorporate cross-hybridization into our statistical model. We show that empirically fitted prior distributions for the cross-hybridization intensities provide sufficient constraints on the model to allow probe-specific correction for cross-hybridization effects, and that the performance of the method is not sensitive to the detailed specification of the prior. Using the Affymetrix U133 human genome spike-in data and the Gene Logic mixture study data as our validation dataset, we compare the expression indexes computed using the random effect model method to several existing methods: two MBEI models proposed by Li and Wong, the Affymetrix MAS 5.0 suite, and the RMA model proposed by Irizarry et al. We found that the random effect approach offers significant improvement over other methods in the following measures: linearity of estimated indexes versus true expression level, measured by both Pearson correlation coefficients and Cosine correlation coefficients; sensitivity and specificity in the detection of differential expression, for both the whole set of genes on the array and a subset of genes detected to be present. We have implemented this approach as a new option in the dChip software (freely available at www.dchip.org). (Abstract shortened by UMI.)
590 $a School code: 0084.
650 4 $a Biology, Biostatistics. $3 1018416
690 $a 0308
710 2 0 $a Harvard University. $3 528741
773 0 $t Dissertation Abstracts International $g 65-10B.
790 1 0 $a Wong, Wing Hung, $e advisor
790 $a 0084
791 $a Ph.D.
792 $a 2004
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3149541