Hi Jiangtao, We do not assign any of the positions for the SNPs shown in the Genome Browser; we take the positions directly from dbSNP. See the "All SNPs(132)" track description page for more information. To get to the track description page, go to the hg19 browser (http://genome.ucsc.edu/cgi-bin/hgTracks?org=Human&db=hg19), scroll down to "Variation and Repeats" and click the blue track name. Under "Data Sources and Methods" you will see exactly where the data comes from.
I do see that the chr1 position is not displayed on dbSNP's page for rs3115865. I am less familiar with dbSNP's display conventions than with the Genome Browser display, and I am not sure why they don't display the chr1 location for that SNP. In the Genome Browser, rs3115865 is not displayed by default in the "All SNPs" track, because by default a filter is applied so that SNPs with a weight greater than 1 are not displayed (this can be changed in the track controls, however). I suggest contacting dbSNP about your questions below, as well as checking out the SNP FAQ Archive at dbSNP for more information on how to interpret dbSNP's data: http://www.ncbi.nlm.nih.gov/books/NBK3848/ -- Brooke Rhead UCSC Genome Bioinformatics Group On 05/28/11 17:26, Tao Jiang wrote: > Dear Rhead, > > For the first example, what record should I keep? > when I look the rsID 3115865 in dbSNP page(http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=3115865), I cannot find the chromosome position in the dbSNP page. Why you assign 741579 on chr1 for this rsID? While for the second rsID, I found the right chr position for it. > > Did you have a detailed illustration on the UCSC FAQ website showing why you keep two(sometimes three or more) rsID for the identical chr position? > > For example, for the three records for one chr position, which should > I keep, the first or the last one, or none of them is a confident SNP??? > > 585 chr1 12479 12480 rs62028715 0 + C C > C/G genomic single unknown 0 0 unknown exact 3 > MultipleAlignments BCMHGSC_JDW, 0 > 585 chr1 12479 12480 rs78725323 0 - C C > C/G genomic single unknown 0 0 unknown exact 3 > MultipleAlignments ENSEMBL, 1 C, 2.000000, > 1.000000, maf-5-some-pop,maf-5-all-pops > 585 chr1 12479 12480 rs78735090 0 + C C > C/G genomic single by-frequency 0.48 0.09798 unknown exact 3 > MultipleAlignments 1 ENSEMBL, 2 G,C, > 4.000000,6.000000, 0.400000,0.600000, maf-5-some-pop,maf-5-all-pops > > But for the SNP on the following chr position: > 585 chr1 10439 10440 rs112155239 0 + C C > A/C genomic single unknown 0 0 unknown exact 1 > 1 BCM-HGSC-SUB, 0 > 585 chr1 10439 10440 rs112766696 0 + C C > -/C genomic deletion unknown 0 0 unknown exact 1 > 1 BUSHMAN, 2 -,C, 1.000000,1.000000, > 0.500000,0.500000, > > which record should I regard as confident? for they both have no "MultipleAlignments" annotation. > > I am eagerly to know all the details and which record(or rsID) should > I consider as confident. > > > All my best, > > Jiangtao > > BGI-Shenzhen > > > On May 28, 2011, at 9:01 AM, Brooke Rhead wrote: > >> Hi Jiangtao, >> >> The engineer who works on SNPs here took a quick look at your examples. For >> the first one, there may have been a strand error that caused dbSNP to fail >> to cluster the newer SNP (rs61770168) together with the older SNP >> (rs3115865). For the second example, you may just want to ignore >> rs28658281, since it aligns to multiple places in the genome and has been >> given a higher weight by dbSNP. >> >> In both cases, you should check out dbSNP's pages (which are linked on the >> details pages in the Genome Browser). For instance: >> http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=3115865 >> >> There you can see the full report on each reference SNP (which start with an >> "rs" -- we only show reference SNPs in the Genome Browser), and you can look >> at the list of submitter SNPs (which start with "ss") that were used to >> determine each reference SNP. >> >> You might also want to report these cases directly to dbSNP at >> [email protected]. >> >> I hope this is helpful. If you have further questions for UCSC, please feel >> free to contact us again at [email protected]. >> >> -- >> Brooke Rhead >> UCSC Genome Bioinformatics Group >> >> >> >> On 05/26/11 21:58, Tao Jiang wrote: >>> Dear UCSC bioinformatic scientists: >>> When use the file "snp132.txt.gz" downloaded from UCSC, I found a confusion >>> question: Why the SNPs that have exactly the same start and end coordinates >>> have different rsID, and sometimes have different allele frequency? Which >>> record is the refSNP record from NCBI dbSNP 132? >>> The following are two examples: >>> 1. 590 chr1 741578 741579 rs3115865 0 - T >>> T A/G genomic single by-cluster,by-hapmap 0 0 >>> unknown exact 3 590 chr1 741578 741579 rs61770168 0 >>> + T T C/T genomic single unknown 0 0 >>> unknown exact 1 1 2. 707 chr22 16050993 >>> 16050994 rs7288968 0 + G G C/G >>> genomic single by-cluster,by-2hit-2allele,by-hapmap 0 0 >>> unknown exact 1 2 BCM_SSAHASNP,CSHL-HAPMAP, 0 >>> >>> 707 chr22 16050993 16050994 rs28658281 0 - >>> G G C/G genomic single by-cluster,by-frequency 0.375 >>> 0.216506 unknownexact 3 MultipleAlignments 6 >>> BCM-HGSC-SUB,BCMHGSC_JDW,BGI,ENSEMBL,HGSV,SSAHASNP, 2 G,C, >>> 6.000000,2.000000, 0.750000,0.250000, >>> maf-5-some-pop,maf-5-all-pops >>> I am looking forward to hearing from you as soon as possible. Thanks a lot! >>> All my best, >>> Jiangtao >>> BGI-Shenzhen >>> _______________________________________________ >>> Genome maillist - [email protected] >>> https://lists.soe.ucsc.edu/mailman/listinfo/genome > _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
