Hello:
I have performed a ChIP-seq experiment with sequencing on SOLiD. My collaborator mapped results with BWA and peak calls with MACS. MACS called peaks but I noticed that the number of tags in our input was much greater (40 to 70 fold) than our IP experiment. For example, for one experiment, total number in tags in input was 52,116,426 (49,941,198 after filtering) and in IP total tags = 1,609112 (713,397 after filtering--a 56% redundancy rate). We see the same thing in a biological replicate and in a separate experiment using the same IP conditions. It was a C.elegans whole animal ChIP experiment.
This difference seems greater than what I have seen in other studies. The samples passed QC at the sequencing center and there were no issues with library construction. For both input and IP, the same amount of DNA was used for library construction.
What could be the source of the difference? Or is this type of discrepancy not unusual? Does it affect how we view the results?
Thanks for any help,
Chris
I have performed a ChIP-seq experiment with sequencing on SOLiD. My collaborator mapped results with BWA and peak calls with MACS. MACS called peaks but I noticed that the number of tags in our input was much greater (40 to 70 fold) than our IP experiment. For example, for one experiment, total number in tags in input was 52,116,426 (49,941,198 after filtering) and in IP total tags = 1,609112 (713,397 after filtering--a 56% redundancy rate). We see the same thing in a biological replicate and in a separate experiment using the same IP conditions. It was a C.elegans whole animal ChIP experiment.
This difference seems greater than what I have seen in other studies. The samples passed QC at the sequencing center and there were no issues with library construction. For both input and IP, the same amount of DNA was used for library construction.
What could be the source of the difference? Or is this type of discrepancy not unusual? Does it affect how we view the results?
Thanks for any help,
Chris