GEO Accession viewer

NCBI > GEO > Accession Display

Not logged in | Login

GEO help: Mouse over screen elements for information.

Sample GSM1479937

Query DataSets for GSM1479937

Status

Public on Oct 20, 2014

Title

H3K27me3 Mel Rep2

Sample type

SRA

Source name

D. Melanogaster_H3K27me3

Organism

Drosophila melanogaster

Characteristics

genotype/variation: wild type
tissue: embryos
age: 4-12 hrs
chip antibody: H3K27me3, Millipore, #07-449

Extracted molecule

genomic DNA

Extraction protocol

Chromatin prepared from Drosophila embryos was immunoprecipitated with specific antibodies and precipitated DNA was isolated (see mat and methods for details)
Libraries were prepared according to Illumina's instructions

Library strategy

ChIP-Seq

Library source

genomic

Library selection

ChIP

Instrument model

Illumina Genome Analyzer IIx

Description

processed data file: dMel_binned_profiles.txt

Data processing

Mapping - For each ChIP-seq experiment, we mapped short reads (36/50 bps) to the appropriate reference genome using Bowtie-0.12.2 by specifying ‘-a –m 1 -best -strata’ options. Our reference genomes included the heterochromatic fraction of chromosomes and non-assembled contigs, and any read that could be mapped to more than one genomic locus was filtered from further processing.
Pile-up - Mapped tags were extended to 140bp and coverage statistics for 20bp bins over each of the genomes were computed. This was done following analysis of the correlation between coverage in the forward and reverse strands, which indicated peak correlation between the two at an offset of 140bp, suggesting the empirical fragment length assayed by ChIP-seq was 140bp.
Normalizing per-species' mappability - Each species genome contains regions that are non-unique and therefore cannot be profiled using short read tags. We generated a control profile for each species by applying the mapping and binning procedure as described above for all possible 36/50bp subsequences in each of the genomes. We derived mappability tracks for each species, indicating the fraction of fragments covering each 20bp genomic bin that could be mapped uniquely to the respective genome. We then normalized the coverage score of each experiment by dividing it with this mapability value. Regions with mappability lower than 50% were filtered out and marked as Not Mapable in subsequent analysis. Note that unmappable regions were different between the species due to duplications and deletions.
Coverage normalization - To control for ChIP-seq coverage and variable ChIP-seq specificity, we transformed raw coverage values to minus log2 of its (1 - quantile) value For example, a normalize value of 9 indicates coverage is in the top 1-2-9 quantile (e.g. – in the top (1/512)th of the distribution).
Genome_build: dm3,droSim1,droYak2,dp4,droVir3
Supplementary_files_format_and_content: For each species we have a tab-delimited file. Each row represents a position in the appropriate genome (1 value for each 20bp bin). The first and second columns are chromosome and position respectively. Columns 3 on represent different ChIP-Seq experiments. Values represent normalized coverage as described in the data processing step.

Submission date

Aug 15, 2014

Last update date

May 15, 2019

Contact name

Bernd Schuettengruber

E-mail(s)

bernd.schuttengruber@igh.cnrs.fr

Organization name

Institute de Génétique Humaine

Lab

Cavalli