Latest revision as of 07:55, 2 July 2010

What I did for today:

Prieto C, Risueño A, Fontanillo C, De Las Rivas J (2008) Human Gene Coexpression Landscape: Confident Network Derived from Tissue Transcriptomic Profiles. doi:10.1371/journal.pone.0003911

"Despite the described interest, coexpression studies done at global “omic” scale are not focused in many cases on human samples [5], and, when they correspond to human, very often they include heterogeneous datasets, mixing “normal” samples with “disease altered” samples from patients suffering from some kind of pathological state. This is the case, for example, in several human gene expression large studies [2], [6]. The inclusion of many disease datasets (mainly from cancer) in such meta-analyses may introduce strong bias and produce a lot of biological noise in the results. In fact, it is well known that cancer cells have altered genomes. Therefore, these kind of studies cannot be used to clarify how a normal-healthy human cellular system works, and they cannot be used to draw a reliable map of the human gene coexpression landscape."

"it has been shown that the inclusion of at least one non-parametric step based on ranks in the analyses of microarray data offers statistically more robust and more accurate estimation of expression values and expression correlations"

"It is difficult to achieve a proper gene coexpression study due to several obstacles that have to be taken in consideration: (i) the technical noise present in the microarrays at genomic scale (ii) the small number of samples used to define each gene expression signature (specially in comparison to the large number of genes); (iii) the strong heterogeneity of the data sets frequently studied, that include in many cases samples from pathological or altered states, which are not adequate samples to find “normal” gene expression behavior." (continue from refs. here....)

TODO for tomorrow:

continue from article above.
study reliability on microarray
http://www.ncbi.nlm.nih.gov/pubmed/12538238
http://www.ncbi.nlm.nih.gov/pubmed/17646307
http://www.ncbi.nlm.nih.gov/pubmed/19212713
http://www.ncbi.nlm.nih.gov/pubmed/17553856
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0003911
http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000807?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+ploscompbiol/NewArticles+(PLoS+Computational+Biology:+New+Articles)

@@ Line 1: / Line 1: @@
-"Despite the described interest, coexpression studies done at global “omic” scale are not focused in many cases on human samples [5], and, when they correspond to human, very often they include heterogeneous datasets, mixing “normal” samples with “disease altered” samples from patients suffering from some kind of pathological state. This is the case, for example, in several human gene expression large studies [2], [6]. The inclusion of many disease datasets (mainly from cancer) in such meta-analyses may introduce strong bias and produce a lot of biological noise in the results. In fact, it is well known that cancer cells have altered genomes. Therefore, these kind of studies cannot be used to clarify how a normal-healthy human cellular system works, and they cannot be used to draw a reliable map of the human gene coexpression landscape." <ref> doi:10.1371/journal.pone.0003911</ref>
+{{Stockholm/Top2}}
-=Reference=
+__TOC__
-{{Reflist}}
+= What I did for today: =
+#'''Prieto C, Risueño A, Fontanillo C, De Las Rivas J (2008) Human Gene Coexpression Landscape: Confident Network Derived from Tissue Transcriptomic Profiles. doi:10.1371/journal.pone.0003911'''
+::* ''"Despite the described interest, coexpression studies done at global “omic” scale are not focused in many cases on human samples [5], and, when they correspond to human, very often they include heterogeneous datasets, mixing “normal” samples with “disease altered” samples from patients suffering from some kind of pathological state. This is the case, for example, in several human gene expression large studies [2], [6]. The inclusion of many disease datasets (mainly from cancer) in such meta-analyses may introduce strong bias and produce a lot of biological noise in the results. In fact, it is well known that cancer cells have altered genomes. Therefore, these kind of studies cannot be used to clarify how a normal-healthy human cellular system works, and they cannot be used to draw a reliable map of the human gene coexpression landscape."''
+::* ''"it has been shown that the inclusion of at least one non-parametric step based on ranks in the analyses of microarray data offers statistically more robust and more accurate estimation of expression values and expression correlations"''
+::* ''"It is difficult to achieve a proper gene coexpression study due to several obstacles that have to be taken in consideration: (i) the technical noise present in the microarrays at genomic scale (ii) the small number of samples used to define each gene expression signature (specially in comparison to the large number of genes); (iii) the strong heterogeneity of the data sets frequently studied, that include in many cases samples from pathological or altered states, which are not adequate samples to find “normal” gene expression behavior."'' (continue from refs. here....)
+== TODO for tomorrow: ==
+*continue from article above.
+*study reliability on microarray
+*http://www.ncbi.nlm.nih.gov/pubmed/12538238
+*http://www.ncbi.nlm.nih.gov/pubmed/17646307
+*http://www.ncbi.nlm.nih.gov/pubmed/19212713
+*http://www.ncbi.nlm.nih.gov/pubmed/17553856
+*http://www.plosone.org/article/info:doi/10.1371/journal.pone.0003911
+*http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000807?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+ploscompbiol/NewArticles+(PLoS+Computational+Biology:+New+Articles)

User:HassanForoughiAsl/1 July 2010

From 2010.igem.org

Latest revision as of 07:55, 2 July 2010

Contents

What I did for today:

TODO for tomorrow: