Chip-seq practical session

Public data, Mikkelsen et al.

Fetch fastq

Go on NCBI GEO

Then select RunSelector.

Chose samples as Lasses' instructions

From a Cell paper from 2010 and belong to this dataset:

http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE20752

Specifically I would look at these samples:

3T3L1_t2_H3K4me3 http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM535748

3T3L1_t3_H3K4me3 http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM535755

3T3L1_t2_H3K27ac http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM535751

3T3L1_t3_H3K27ac http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM535758

I think this is the input control they used for normalization: http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM535740

Download SRR_Acc_List.txt in /work/users/aginolhac/chip-seq/doctoral_school/Mikkelsen

then fetch and compress

parallel -j 6 --progress "fastq-dump --gzip {}" :::: SRR_Acc_List.txt

mapped with paleomix

paleomix bam_pipeline --bwa-max-threads=4 --max-threads=12 mikkelsen.yml

last for ~ 3 hours 30 minutes

peak calling

H3K4

macs2 callpeak -tMikkelsen_3T3L1_t2_H3K4me3.GRCm38.p3.bam \
               -c Mikkelsen_3T3L1_WCE.GRCm38.p3.bam \
               -f BAM -g mm -n 3T3L1_t2_H3K4 -B -q 0.01 --outdir 3T3L1_t2_H3K4 &
macs2 callpeak -t Mikkelsen_3T3L1_t3_H3K4me3.GRCm38.p3.bam \
               -c Mikkelsen_3T3L1_WCE.GRCm38.p3.bam \
               -f BAM -g mm -n3T3L1_t3_H3K4 -B -q 0.01 --outdir 3T3L1_t3_H3K4

H3K27

macs2 callpeak -tMikkelsen_3T3L1_t2_H3K27ac.GRCm38.p3.bam \
               -c Mikkelsen_3T3L1_WCE.GRCm38.p3.bam --broad  \
               -f BAM -g mm -n 3T3L1_t2_H3K27ac -B -q 0.01 --outdir 3T3L1_t2_H3K27ac &
macs2 callpeak -t Mikkelsen_3T3L1_t3_H3K27ac.GRCm38.p3.bam \
               -c Mikkelsen_3T3L1_WCE.GRCm38.p3.bam --broad \
               -f BAM -g mm -n3T3L1_t3_H3K27ac -B -q 0.01 --outdir 3T3L1_t3_H3K27ac