Why is the cds start and cds end position the same in some of the known-gene data downloaded from ucsc ? (ucsc table browser)

Question

[mlank001]

I am working on human cancer data and wanted to extract sequences from the human genome for analyzing SNPs. For that, I am trying to fetch exon start and end positions for the human genome build GRCh37. I used ucsc genome browser to download this data (using known Gene table)

The data has the same values for cds start and cds end. I am confused as to why this is the case, as cds is for coding region. Should I ignore this and go use exon start and end positions?

Any help is highly appreciated!

Thanks!

[PoGibas]

Please post example of "The data has the same values for cds start and cds end".

[mlank001]

Here is the example of one entry:

#na           chrom    strand    txStart    txEnd    cdsStart    cdsEnd    exonCount    exonStarts            exonEnds              proteinID      alignID
uc001aaa.3    chr1     +         11873      14409    11873       11873     3            11873,12612,13220,    12227,12721,14409,    uc001aaa.3

If you look at the 6th and 7th columns (cdsStart and cdsEnd), the values are the same. I am confused as to why that's the case.

Answer 1

[PoGibas]

Answer: This gene doesn't have normal usual structure as it is a long non-coding RNA (transcribed_unprocessed_pseudogene).

Explanation: In UCSC browser find gene name: DDX11L1.

curl -s "ftp://ftp.sanger.ac.uk/pub/gencode/Gencode_human/release_21/gencode.v21.annotation.gtf.gz" |   
    gunzip -c | 
    tr -d ';"' | 
    awk '($3=="gene" && $18=="DDX11L1") {print $14}'

transcribed_unprocessed_pseudogene

[mlank001]

Got it! thanks a lot! I appreciate it..

So this means I should indeed be looking at the cds Start and end positions for extracting the reference genome sequence to be translated.

[PoGibas]

It depends. Are you interested only in protein coding genes? If so, check genecode genes, you can download list of all the protein coding genes (start end positions for gene/transcript/UTR/exon/CDS and genomic or translated fasta sequences).

Answer 2

[Chirag Nepal]

In general, CDS start-end, they are noncoding transcripts