Hello! For the project I am working on, I need to design CRISPRi guides against ZNF217. I am running into some issues & was wondering if anyone could help!
Firstly, I cannot pin down the correct annotation for the TSS of this gene. Eukaryotic Promoter Database (EPD) indicates two different, experimentally-determined promoters, which I guess correspond to two different transcripts. However, when using ZNF217 gene ID to design CRISPRi guides using the Broad Institute guide design tool, their TSS annotation does not fall at either of these promoters, but in between them (~11 kb away from either one, but lines up with the largest ATAC-seq peak close to ZNF217 that certainly LOOKS like it could be a promoter). I am wondering if there’s any way to figure out what is the correct/most likely annotation such that I can knockdown this gene, without having to target each of these locations?
From top to bottom: (1) combined ATAC peaks from stomach cancer cell lines, (2) ZNF217 promoters from Eukaryotic Promoter Database, (3) ZNF217 TSS according to Broad institute guide design tool, & (4) ZNF217 transcript location according to Refseq genes. This is all in hg19, btw.
Secondly, I designed my guides using the Broad Institute tool; however, I thought I would try to "check" some of them using IDT’s gRNA checker tool & was a surprised to find that the predicted on- & off-target efficacy scores were WAY WAY off between the two. Now, I don’t know mathematically how either of these tools come up with their efficacy scores, so I’m not looking for them to be identical, but all the top guides from Broad’s tool looked pretty abysmal according to IDT. Am I missing something? Can I not do this?
An example: This guide (the top-ranked guide acquired from Broad tool using the following settings: CRISPRi, spyoCas9, ZNF217 gene ID as target) has an on-target efficacy score of 0.6888 for Broad tool (which should be a good score!) & no indication of "other target matches." Using the IDT tool, I see an on-target score of 12 (really a very bad score!). Also confused that the 100% sequence match on another chromosome does not affect the off-target score.
tl;dr: I am very confused as to what I can do to design guides that are likely to work for me or how to understand these tools better. I have already experimentally tested two guides which did not knockdown my gene & would really like to get it right this time. If you have any experience or insight regarding CRISPRi guide design in general, understanding the differences between the Broad & IDT tools, or how to navigate confusing genome annotations, I would really appreciate some help as I don’t have anyone I can really go to in my lab. :(
