I have been wondering why many around the world are re-processing some (identical) fraction of TCGA data again (as far as simple analyses are concerned e.g. RNAseq)? Surely these types of analyses have already been done (using a reasonable set of parameters/programs) and the result files are potentially publicly available.
I think there are several possible reasons:
1, Different demand: sometimes I need to analysis the RNA-seq and methylation data simultaneously with same annotation. Therefore, sometime, I need to download the Raw Data and try to analysis them again with my own frame.
2, Different analysis method: Dozens of analysis methods to same dataset, such as Bismark, BSMAP and so on. In addition, different normalization way, different differential statistic test method, different missing value imputation way. Anyway, You have too many choice to select different pre-processing treatment.
3, Public data and local data integrative analysis: sometimes, I have my own data and public data, I want to integrate them. In this situation, analyze them with same pipeline would be more acceptable.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
These scenarios are specialized circumstances and are certainly applicable in some cases. I find it hard to believe that everyone trying to use TCGA data has a real need to download and re-process data.
I wonder how many users do it because they are not able to find processed data or know that they can use results from one of the available portals that provide access to derived data.