I am analysing chip-seq data of RNA Pol-2 generated in our lab. But I am not getting the desired peaks on genes. Can somebody help? I am attaching some IGV screenshots for your reference. The Input has 22 million reads and the RNA Pol-2 chip has 23 million reads. I used bamcoverage command to convert the bam files to bigwig files.
1 answer
I have no good answer to that. I guess DNA recovery relative to IgG gives some hints. My old PI always insisted on qPCR but depending on choice of primers that was more of a gamble than actual quality assessment. What I would probably do is to try some protocols, some antibodies and make the libraries, then squence shallowly. You will see even in shwllow-seq if things are just noise or actual peaks. I am not much into Pol (it is more a strech of signal, is it?) so I would get published data, see how these look, and then maybe subsample their fastq files, checking what is the minimum depth necessary for a good dataset to see some peaks. That is then the target depth for your shallow exploration panel.
I would also say that the ChIP failed.
Before doing an ChIP with an unkown antibody, apart from the datasheet of the company declaring that this is ChIP specific, I would have performed a western blot and immunofluorescence.
which washing solutions you used and which protocol you followed. What was your crosslinking strategy? PolII antibodies are relatively good and PolII is nicely bound to chromatin, so you should do harser washing.
Also, a good positive control is the H3 antibody. needs few reads.
Before, committing to sequencing, ask for the NGS center to do a "shallow sequencing" that means 10x or even more less reads that the wanted reads. This will provide a very good overview of the quality of the ChIP.
Computationaly the only , as far as I know, good indication of a working ChIP is calculating the FRiP, for both the Antibody and the IgG.
Please also note that the reads the IgG are the same as with the IP. You would expect at least twice that, to get a good signal.
Image of how it should look
Log in to answer this question.
-
Generate density plot for methylation data
written by ty.fu •
-
Human sequence data mapped to mouse genome
written by sh •
-
Wrong in Deeptools : the enrichment of TSS and TES region are similar when using scale-region
written by hao •
-
Error while running Virusfinder
written by DC •
-
Long read RNA # of reads interpretation help
written by shinyjj •
-
iTOL phylogenetic tree visualization
written by Confused_human •
-
Seurat subset cluster cell annotation
written by leranwangcs •
-
SNP Pruning Through PCA (Edit: Feature Selection Through PCA)
written by ErickW •
-
pycoQC Report - Interval unit in output over time plot
written by chrys •
-
mapped reads (m aligned) is higher for a subset sample?
written by Pratik
