wiki:BIOS_PreparedData

Version 1 (modified by Rick, 9 months ago) (diff)

--

Recommended BIOS datasets for downstream analysis

RNAseq data

Freeze I

Data available

Raw RNA seq data is avalable at the grid, see RNASeq data. This data has been aligned using the pipeline described at RNAseq alignment and quantification pipeline, the exon, transcript and gene level count output is described in the following. Count data is available from the so called 'Freeze1': These are the 2116 samples from Groningen (N=626), Leiden (N=654), Rotterdam (N=652) and Maastricht (N=184) that passed QC (see RNAseq QC). This is around half of the BIOS RNA seq data that is used for the first papers: the other half has been measured but is still in the process of aligning and QC. Both raw and TMM normalized data are available. TMM normalization corrects for the different library sizes across subjects, see attached script for R code or the R package edgeR, and http://genomebiology.com/2010/11/3/r25.

Location on VM

  • exon counts raw: /virdir/Backup/RP3_data/RNASeq/run_01/exoncounts/exon_count_freeze1_R_object.RData
  • exon counts TMM normalized: /virdir/Backup/RP3_data/RNASeq/run_01/exoncounts/exon_count_freeze1_TMM_normalized_R_object.RData
  • transcript counts raw: /virdir/Backup/RP3_data/RNASeq/run_01/transcriptcounts/transcript_count_freeze1_R_object.RData
  • transcript counts TMM normalized: /virdir/Backup/RP3_data/RNASeq/run_01/transcriptcounts/transcript_count_freeze1_TMM_normalized_R_object.RData
  • gene counts raw: /virdir/Backup/RP3_data/RNASeq/run_01/genecounts/gene_count_freeze1_R_object.RData
  • gene counts TMM normalized: /virdir/Backup/RP3_data/RNASeq/run_01/genecounts/gene_count_freeze1_TMM_normalized_R_object.RData

How to use the data

Data is stored in R objects, by loading the files in R (e.g. type load('/virdir/Backup/RP3_data/RNASeq/run_01/exoncounts/exon_count_freeze1_R_object.RData') in R) there will be a matrix called RNA in your workspace for the raw data, and RNAs for the TMM normalized data. The row names of these matrices (type rownames(RNAs)) contain gene, exon or transcript IDs, the column names (colnames(RNAs) are the subject BIOS IDs (called uuid in other files). We used ensembl v.71 for annotation, see Reference and annotation. If you want to export the data to a tab delimited text file, use write.table(RNAs, file='yourfile.txt', quote =FALSE, col.names=TRUE, row.names=TRUE, sep='\t').

Freeze II

Data available

Location on VM

How to use the data

DNA methylation data

Data available

Raw methylation data, idat-files, are available for all samples currently (01-08-2015) generated within the BIOS project. Sample swaps detected by MixupMapper (http://bioinformatics.oxfordjournals.org/content/27/15/2104) have been corrected. Detection of bad quality samples or runs can be performed using MethylAid (http://bioinformatics.oxfordjournals.org/content/30/23/3435) the results are available at http://shiny.bioexp.nl/BIOS. Rdata-files have been generated for each biobank and a combined file for all biobanks containing SummarizedData-objects with functional normalized beta values sample and feature annotation. QC has been perform using MethylAid.

Location on VM

The idat-files are located on: /virdir/Scratch/RP3_data/IlluminaHumanMethylation450k/450k separated by biobank: /virdir/Scratch/RP3_data/IlluminaHumanMethylation450k/450k/CODAM, /virdir/Scratch/RP3_data/IlluminaHumanMethylation450k/450k/LL and all folders contain the samplesheets provided by the data generation center including a unique and universal BIOS identifier (uuid in phenotype files).

How to use the data

The Bioconductor/R packages minfi and illuminaio provide reading capabilities for the idat-files.

Genotype data

Data available

All genotype data that has been imputed(GoNL version 5) and stored on the SRM is also available from the VM. Beware the imputed files contain indels as well and the Biobank person_id is used in the sample-files.

Genotype data has also been imputed using the Human Reference Consortium (HRC) reference panel and is available from both the SRM and the VM. The same warnings apply for this set.

Location on VM

Gzipped IMPUTE2 files are stored at /virdir/Backup/RP3_data/GWAS_ImputationGoNLv5 per biobank.

HRC imputed data is contained in per-chromosome Variant Call Format (VCF) files at /virdir/Backup/RP3_data/HRC_Imputation/[Biobank]/results/unzipped/..

How to use the data

Once unzipped IMPUTE2 files can easily be read as these are tab-separated files (see https://mathgen.stats.ox.ac.uk/impute/impute_v2.html).

Note that the HRC imputed data is in VCF format, which you may need to convert before usage.

Phenotype data

Data available

BIOS phenotype data is stored in a meta database, see this page. This databases can be accessed by so called views, using e.g. R. Three views were extracted (January 2015) and stored at the VM:

  • phenotype data: view="allPhenotypes" design="phenotypes"
  • RNA seq sample sheets: view="rnaseq", design="samplesheets"
  • IDs: view="getIds", design="identifiers"

These files are available in .RData and .csv file formats. See for column name explanations the page Phenotype data. Phenotype data is not complete yet: we are currently contacting the biobanks to complete there files.

Location on VM

  • Phenotypes: /virdir/Backup/RP3_data/Phenotypes/BIOS_Phenotypes.RData (matrix P) and /virdir/Backup/RP3_data/Phenotypes/BIOS_Phenotypes.csv
  • RNA-seq sample sheets: /virdir/Backup/RP3_data/Phenotypes/rna_seq_sample_sheets.RData (matrix S) and /virdir/Backup/RP3_data/Phenotypes/rna_seq_sample_sheets.csv
  • IDs: /virdir/Backup/RP3_data/Phenotypes/BIOS_IDs.RData (matrix F) and /virdir/Backup/RP3_data/Phenotypes/BIOS_IDs.csv

How to use the data

Link the files to the RNA-seq, genotype or methylation data by mapping the corresponding IDs.

Attachments (1)

Download all attachments as: .zip