Loading...
 

ChIP-Seq Analysis Using MACS and MEME

1. Peak Finding

1.1 Example Data

a. Login to MolBioCloud using X2GO.

b. Create a ‘macs2’ folder in your Desktop.

c. Download the MACS example dataset for FoxA1 and save it to the macs2 folder.

d. The example dataset consists of Input_tags.bed and Treatment_tags.bed for FoxA1 treatment.

1.2. Start the terminal

Start the terminal and change to the macs2 directory

1.3. Run macs2

Run macs2 using the following commands;

macs2 callpeak -t Treatment_tags.bed -c Input_tags.bed -f BED -g hs -n FoxA1 -B –p 0.005

The –t option specifies the treatment bed file, -c specifies the input (control) bed file, -g specifies the human as the genome, -n specifies the name for the output file, -B specifies that a bedgraph pileup is stored in the current directory and –p specifies the pvalue cutoff threshold.

 Running TIme

It takes about 5 minutes or so to run this data set.

1.4. Output files

FoxA1_peaks.narrowPeak

FoxA1_peaks.xls

FoxA1_summits.bed

FoxA1test_model.r

FoxA1_control_lambda.bdg

FoxA1_treat_pileup.bdg

Please see the MACS site for more info on passing different other parameters.

https://github.com/taoliu/MACS

1.5. Model PDF

Run ‘Rscript FoxA1_model.r’  in the shell to generate the PDF of the model


2. Extract Peak Sequences

Run the below commands to get some idea about your results

wc -l Input_tags.bed
wc -l FoxA1_summits.bed
grep -w 'chr1' FoxA1_summits.bed | wc -l

2.1. Extract top 1000 peaks

...from the FoxA1_summits.bed using the below command;

head –n 1000 FoxA1_summits.bed > 1000summits.bed

2.2. Reformat BED file

Convert the 5 column 1000summits.bed to 4 column 1000summits4.bed file by cutting and saving only the first 4 columns;

cut –f 1-4 1000summits.bed > 1000summits4.bed

2.3. Extract FASTA sequences

Extract FASTA format sequences of the 1000 peaks, using UCSC genome browser to table browser option, as a custom track, save it as sequence.fasta.


3. MEME-CHIP

Run meme-chip for two motifs, choosing DNA as molecule with ZOOPS option.

meme-chip -meme-mod zoops -meme-nmotifs 2 1000summitsequence.fasta

4. Peak Annotation

Run chipseek with 4 column data