## page was renamed from CancerStemCellProject/ChangjiangXu/BiostatService/Expectations
##master-page:HomepageTemplate
#format wiki
#acl All:read

== Expectations ==
##Expectations
##Expectations from Statistical Consulting/Analysis of Gene Expression Experiments

Statistical consulting analysis support for experiments in the Cancer Stem Cell Program is available by contacting Biostatistician Changjiang Xu at changjiang.xu@utoronto.ca.  The mandate is to do both statistical analysis and education of trainees/investigators with respect to statistical issues and concerns.

'''Statistical analysis is available for any problem related to the cancer stem cell network:'''
 * Statistical analysis of gene expression experiments
 * Design of analysis of experiments to improve efficiency and aid in detection of desired results.
 * Multiple testing adjustments and calculation of false discovery rates.
 * Meta-analysis of multiple studies
 * Survival analysis, multivariate analysis, etc, and other problems as needed

'''The following deliverables can be expected for the analysis of any study/data set:'''
 * An initial meeting with the Investigator to discuss the goals of the study, the proposed analysis, best design (if the experiment has not yet been done), and issues with respect to design and analysis that are relevant.
 * A rough estimate of the timeline for primary analysis. If the analysis is non-standard or requires additional consultation and thought, this will lead to uncertainty in the timeline, and this will be conveyed to the Investigator.
 * A document describing all files prepared as the result of analysis.
 * A brief summary of the statistical analysis methods used.
 * A summary meeting with the Investigator, if desired, to discuss the methods used and to describe any unique issues.
 * The possibility for subsequent, targeted, additional analyses after examination of the primary results.

'''The following deliverables can be expected specifically when analyzing gene expression studies:'''
 * We provide a list of all probes and the statistical results, ranked with respect to FDR.
 * The list includes columns of log2 based mean expressions in separate treatment groups, log2FC, t/F-statistic, q values
 * With your analyzed data we concatenate the standard annotations.
 * Education with respect to interpreting the results.
 * Some education as to how to work with the Excel files of results, if needed.

'''Statistical analysis does not include the following services:'''
 * We do not keep backups of your original data files.  Please ensure that you have appropriate backup before bringing data for analysis.
 * We cannot provide additional annotation information or interpretation.
 * For data interpretation (Pathway and Network Analysis), you can contact Dr. Veronique Voisin (Ph.D, Bioinformatician) at veronique.voisin@gmail.com.
##We assume that you can manipulate your results excel files to extract desired summaries.  This would include simple filtering to extract the most significant results, or simple transformations (from p-value to - log10 p-value).
 * Analysis is usually performed on a log-scale for methodological reasons. Hence we will not provide results back-transformed to the linear scale.