Xin,

You will need to change the penalty weight from 

the default

weights = c(0, 0, 0.014, 0, 0, 0.395, 0.317, 0, 0.389, 0.079, 

        0.445, 0.508, 0.613, 0.851, 0.732, 0.828, 0.615, 0.804, 

        0.685, 0.583)

To 

weights = c(0,0,0,0,0, 0, 0.014, 0, 0, 0.395, 0.317, 0, 0.389, 0.079, 

        0.445, 0.508, 0.613, 0.851, 0.732, 0.828, 0.615, 0.804, 

        0.685, 0.583)

Please feel free to modify the 4 zeros (colored in yellow) to any number between 0 to 1 to the best of your knowledge.

Best,

Julie

From: "Gao, Xin (Daniel)" <Xin.Gao@umassmed.edu>
Date: Friday, July 17, 2015 1:17 PM
To: Lihua Julie Zhu <julie.zhu@umassmed.edu>
Cc: "Sontheimer, Erik" <Erik.Sontheimer@umassmed.edu>
Subject: RE: CRISPRseek to analyze NmCas9

Hi Julie,

Yes, I did. By setting gRNA.size=24, I could obtain gRNAs. However, when I set gRNA.size=24 in offTargetAnalysis to search off target site, it didn't work (it only works at gRNA.size=20, and I am sure there is an offTarget site which meets search criteria but CRISPRseek cannot pick it up). Here is an example:

Results2 <- offTargetAnalysis(inputFilePath, findgRNAsWithREcutOnly = FALSE, findPairedgRNAOnly = FALSE, BSgenomeName = Hsapiens, chromToSearch = "all", gRNA.size=24, PAM = "NNNNGCTT", PAM.size=8, PAM.pattern="NNNNGSTT$", txdb = TxDb.Hsapiens.UCSC.hg19.knownGene, orgAnn = org.Hs.egSYMBOL, max.mismatch = 5, outputDir = outputDir, allowed.mismatch.PAM = 5,overwrite = TRUE)  

>>> Finding all hits in sequence chr1 ...

Done searching

......

Building feature vectors for scoring ... 

Error in buildFeatureVectorForScoring(hits = hits, canonical.PAM = PAM,  : Empty hits!

In addition: Warning message:

In searchHits(gRNAs = gRNAs, PAM = PAM.pattern, BSgenomeName = BSgenomeName,  :

  No matching found, please check your input sequence, and make

            sure you are using the right genome. You can also alter your 

            search criteria such as increasing max.mismatch!

My input sequence is: 

 ACACACACGTCCTCACTCTCGAAGACGCTGCTCGCTCCATTCACCCAGCTTCCCTGTGGTGGCCGAGCGCCCCCTAGTGACTGCCGTCTGCACACCC

 CGGCTCTGGCTAAAGAGGGAATGGGCTTTGGAAAGGGGGTGGGGGGAGTTTGCTCCTGGACCCCCTATTTCTGACCTCCCAAACAGCTACATATTTG

 GGACTGGAGTTGCTTCATGTACAGAGA

CRISPRseek finds one of gRNA is (PAM is underlined)

I expect if I set gRNA.size=24, max.mismatch=4, CRISPRseek at least will show me one off-target site. However, it showed me this warning message. Even when I change max.mismatch to 4 or 5 or 6, allowed.mismatch.PAM to 4 or 5 or 6, PAM.pattern="NNNNGGTT$" or "NNNNGVTT". There is always an warning message. Did I do something wrong or CRISPRseek needs some modification?

Thank you very much if you can answer my question!

Best,

Daniel 

From: Zhu, Lihua (Julie)
Sent: Thursday, July 16, 2015 7:05 AM
To: Gao, Xin (Daniel)
Cc: Sontheimer, Erik
Subject: Re: CRISPRseek to analyze NmCas9
 

Daniel,

Please set gRNA.size=24 in the search and you should get what you need !

Best,

Julie

On Jul 16, 2015, at 1:16 AM, Gao, Xin (Daniel) <Xin.Gao@umassmed.edu> wrote:
 

Hi Julie,

Thank you for your help on updating my CRISPRseek to 1.9.1 last time. After I used offTargetAnalysis to predict off-target sites, I noticed one problem after I checked the results. The CRISPRseek could recognize our Nm PAM which is NNNNGATT, however the length of predicted guide RNA should be 24nt which is different from 20nt of S.pyogenes. Our target site is a sequence of 24nt plus PAM sequence NNNNGATT. I mentioned in my first email but I didn't highlight this difference in our later conversation. I am sorry about the inconvenience. I don't know how difficult you will modify CRISPRseek by incorporating this change. Can you please let me know? Thank you very much!

Regards,

Daniel