User:Diego Forero/Notebook/Silico-Meta/2009/09/18

From OpenWetWare

Jump to: navigation, search
Project name A SURVEY OF AVAILABLE COMPUTATIONAL TOOLS FOR META-ANALYSIS OF MICROARRAY EXPRESSION STUDIES
Next entry

General Things

A SURVEY OF AVAILABLE COMPUTATIONAL TOOLS FOR META-ANALYSIS OF MICROARRAY EXPRESSION STUDIES

I am planning to write a paper describing the comparison of available tools for meta-analysis of microarray studies. This idea appeared as part of the plans for the meta-analysis of microarray data from post-mortem brains for AD and PD.


I searched in PubMed, Google Scholar and in the extended lists of R and Bioconductor for packages for meta-analysis of microarray studies. I found 10 packages: BayesPoolMicro, GeneMeta, MAID, metaArray, metaGEM, metahdep, metaMA, OrderedList, RankAggreg and RankProd. I posted the basic list in a Google Sites webpage: http://sites.google.com/site/compgensite/meta

MAID and metaGEM are not available in R or bioconductor. I have a local copy of metaGEM, previously sent by the author. GeneMeta is not described as a paper. I wrote an email to the authors of MAID and metaGEM to ask for their availability (ramasamy@stats.ox.ac.uk and Ian.McGilvray@uhn.on.ca) and to the authors of GeneMeta about plans for publishing a paper about it. Dr Ramasamy answered informing that he will try to have the package in Bioconductor soon. I received a copy of MAID from the authors.


I wrote a possible title and abstract for the paper

  • A SURVEY OF AVAILABLE COMPUTATIONAL TOOLS FOR META-ANALYSIS OF MICROARRAY EXPRESSION STUDIES
  • Abstract: Meta-analysis of genome-wide expression data generated by microarray experiments has been used as a novel approach to discover the transcriptional alterations underlying complex biological phenotypes. This methodological approach represents a potential strategy to find consistent gene expression signatures in larger datasets and to find possible causes of heterogeneity. A comparison of the existing programs designed for implementation of this type of meta-analyses of high-throughput data is lacking. In this work, we review the features of eight freely available computational tools for meta-analysis of microarray data: GeneMeta, MAID, metaArray, metaGEM, metahdep, metaMA, RankAggreg and RankProd. All of these tools are packages for the R/Bioconductor environment and are based on different, and potentially complementary, algorithms. This survey will be useful for researchers interested in the implementation of meta-analyses of microarray data and for the development of future additional computational tools.

I sent a presubmission enquiry to the editor of 6 journals:

  • Briefings in Functional Genomics and Proteomics (answered and suggested to send it to Briefings in Bioinformatics)
  • Briefings in Bioinformatics (answered positively and created an author profile. Journal selected for submission)
  • Comparative and Functional Genomics (answered positively)
  • BMC Genomics (no answer yet)
  • Computers in Biology and Medicine (answered positively)
  • Computational Biology and Chemistry (answered and suggested to send it to other journal)

Dear Editor, we are working in molecular genetics and genomics of human disorders and we have published a number of papers in this area (Forero DA, J Neural Transm 2006; Forero DA, Neurosci Res 2006; Forero DA, Arch Med Res 2006; Forero DA, J Cell Mol Med 2006; Forero DA, J Psychiatry Neurosci, 2009). We have finished a manuscript describing a comparison of the features of eight available computational tools for meta-analysis of microarray data. We would like to know if this manuscript could be of interest for your journal. We are attaching the abstract at the end of the message. Thanks for your attention. Sincerely. Diego Forero, MD//Neurosciences Research Group//MSc Program in Neurosciences//Institute of Genetics-School of Medicine//National University of Colombia//Bogotá, Colombia//daforerog@unal.edu.co// Editor, hum-molgen.org

  • Answer from Briefings in Bioinformatics

Dear Dr Forero, Thank you for agreeing to submit the manuscript entitled "A SURVEY OF AVAILABLE COMPUTATIONAL TOOLS FOR META-ANALYSIS OF MICROARRAY EXPRESSION STUDIES" to Briefings in Bioinformatics. Please try your best to submit this manuscript within the next 30 days. To access the manuscript, login to Briefings in Bioinformatics - Manuscript Central site at http://mc.manuscriptcentral.com/bib. Once you are logged in, the Main Menu will be displayed. Please click on the Author Center, where you will find the manuscript listed under "Invited Manuscripts." You can click on the "Continue Submission" button to begin manuscript submission. All communications regarding this manuscript are privileged. Any conflict of interest must immediately be reported to me. Thank you for agreeing to submit this manuscript. Yours sincerely, Alison Bentley. Briefings in Bioinformatics Editorial Office. briefings@oxfordjournals.org


  • I selected some relevant references, to import them in Reference Manager (PMIDs):

16982708 17900369 18767902 8616827 19228411 19628502 19648140 19563634 17182626 17921495 17651921 12424109 15876355 19189975 15340489 18992145 17726228 12461517 12454644 11389458 19033363 12855442 18204063 16964229 19174838 16369572 17343745 15693945 15184677 14550631 12154050 18940857 19015125 18629145 18463138 17983263 15114354 17411443 18950506 15952881 19034265 11726920 19117983 16485019 12664684 7569999 15461798 15826355 15327980 16872483


THINGS TO DO:

  • PubMed DONE

search "microarray*"// search "microarray* and meta-analysis"// years 1996-2008

  • ArrayExpress-NCBI GEO DONE

statistics 2003-2009

  • 20 papers about applications of meta-analyisis from PubMed Central:

Use or not of a previously available tool for meta-analysis

  • For each one of the eight tools:

Number of citations (Google Scholar). Number of downloads (Bioconductor). Number of citations for method paper. Type of inputs. Type of methods implemented. Type of outputs. Additional comments.

  • To test in each tool

2 different datasets (3 studies for each one): 10 samples, 1000 probe datapoints.// 20 samples, 1000 probe datapoints.// To describe: Processing time, outputs, consistency between tools.


  • I looked into PubMed Central about papers describing papers of applications of meta-analysis of microarrays:

PMC2582039 PMC2515154 PMC2495035 PMC1458884 PMC1904223 PMC2275244 PMC2743637 PMC1664699 PMC2642844 PMC1180834 PMC2098839 PMC2640456 PMC2444034 PMC1273632 PMC1906587 PMC1636648 PMC1853113 PMC2650420 PMC2716681 PMC2525640 PMC1716187 PMC2633354

Further details (and active links) are given in http://sites.google.com/site/compgensite/meta-applied




Personal tools