Hi Stefano, Sorry for the delayed response. There were a few of us looking at this one.
First, the sequence you were expecting is there, just in a different reading frame. To see, follow these steps: 1. Blat for protein sequence: YECNECGKSFHQKANLQKHQGIHTGEKP 2. Click (browser) on the 100% identity match 3. In the browser, turn the Base Position track to display: full. You'll now see the 3 reading frames shown at top. 4. You'll notice that the Blat result is on the reverse strand ( <<< ), so click the small arrow pointing right at the left of the base position track view so that it reverses/points left - so the base position is also reading in that direction. You'll now see that the top reading frame shown, from right to left, is the sequence: YECNECGKSFHQKANLQKHQGIHTGEKP. One of our engineers added: "What we suspect is that the DNA sequence that codes for this protein in one frame is a motif found in ZNFs. However, this DNA sequence is frame shifted in ZNF717 to code for a different protein. Several sources of evidence indicate that the ZNF717s CDS annotation is most likely correct. Another possibility is that the protein sequence is incorrect in the database, coming from an incorrect computation translation of an mRNA." Please let us know if you have any additional questions: [email protected] - Greg Roe UCSC Genome Bioinformatics Group On 8/1/11 7:40 AM, [email protected] wrote: > Dear All > > I'm Stefano Berto and I'm working in the Katja Nowick group as a PhD > student. In the signature you can easily find my project. > > Nowadays I'm working with Transcription factors and the differences > between humans and closely related species. > Thus I found very useful your accurate annotations. > > Unfortunately I detected an error downloading a sequence of ZNF717. > > This Zinc Finger can be easily found in catalogue: > http://znf.igb.uiuc.edu/human/action/exploreView?type=locus&id=732 > > and also in your database, Blatting the sequences and find the locus in > the Chromosome. > > The problem that I noticed is the protein sequence of ZNF717. > > From > http://znf.igb.uiuc.edu/human/action/exploreView?type=locus&id=732 > I checked the domains of the protein (KRAB, fingers etc) finding a correct > annotation of the protein. > > But if I download the protein sequence there is an error. > In fact, I tried to detect the Zinc Finger domains in the sequence > download and I found some mistakes about some undetectable domains. > > Here I'm explaining the problem: > > http://znf.igb.uiuc.edu/human/action/exploreView?type=locus&id=732 > Here there are the list of domains. > I took this ZNF domain: > YECNECGKSFHQKANLQKHQGIHTGEKP > This is the 18th ZNF domain but I noticed the error in other motifs. > > Now I download the protein sequence from the same catalogue and I tried, > with the alignment tool, to detect the sequence but unfortunately I > couldn't. > > > Thus I blatted the finger sequence in the UCSC website finding a 100% of > identity with a locus in the Chr 3. It is completely inside the ZNF717. > > The ERROR is here. In fact I clicked in the browser and, as you understand > if you are following my steps, the SEQUENCE in the browser is completely > different: > MNVMNVENPFIRRQIFRSIKVFTRGRN > > Moreover I double checked with the translation of the chromosome locus and > it will give me the CORRECT zinc finger sequence > (YECNECGKSFHQKANLQKHQGIHTGEKP). > > Moreover the Refseq in your database is the same that I had from the > catalogue and it is wrong too. > > Thus I guess there is a mismatch between the database and the reference > sequence that you can download. > I got in one hand one plausible result but on the other hand a wrong > reference sequence. > > Could you check the error? > > I hope you can understand the explanation. > I did my best. > By the way, if you need more information, don't hesitate to contact me. > > Looking forward to hear from you soon > Best Regards > Stefano > > _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
