Discrepancy between positions on IGV and NCBI's PrimerBLAST

Question

[Rozita]

Hi,

I've got a set of primers that I want to check if they span regions where my protein of interest is enriched from publicly available datasets, so I created a bed file which contains the positions of the primers (which I've lifted off a paper and verified using PrimerBLAST) and I noticed that the position shown on IGV and PrimerBLAST is different for the same set of primers, even though I'm sure I'm using hg38 as a reference genome.

Any suggestions as to why there is this discrepancy?

Thank you.

[GenoMax]

It is difficult to see what you want to illustrate from the screenshot attached. Can you explain what is fhe difference that you are observing or post a better screenshot.

[Rozita]

On IGV, it shows that the primer is spanning the promoter region, whereas in PrimerBLAST it is showing that it is spanning an intron in the middle of the gene.

This is also the case for other primers.

[GenoMax]

Can you check with the in silico PCR tool at UCSC to see what that says : https://genome.ucsc.edu/cgi-bin/hgPcr

PrimerBLAST is likely accurate though.

[Rozita]

It showed me the same positions as that of PrimerBLAST.

Any idea how do I now visualise the publicly available datasets that I have and compare their positions with that of the regions where my primers are targeted? And any possible explanation why this is happening on IGV? It seems very odd, but I don't understand what could be wrong.

[GenoMax]

It showed me the same positions as that of PrimerBLAST.

So we know that result is correct.

You made the BED file manually? It is in correct format? Is IGV showing that feature in right spot? If not, you will need to make sure if you found a new bug by doing additional testing.

[Rozita]

Yes, I did. I entered the positions of the primers from PrimerBLAST in a CSV file then converted it to a BED file which opens successfully on IGV, when I search for these primers I can see that they show up on the correct chromosome number and within the correct range in terms of position, however, that position doesn't correspond to my gene of interest as I see on PrimerBLAST.

So like if one of my primers on PrimerBLAST has a start position of 5000 and end position of 5020 at chr1 and it shows that it is at the promoter of Gene X, when I load the BED file on IGV and I search for chr1:5000-5020 I can see my primer showing at that region, however, when I check the RefSeq track, it doesn't correspond to Gene X or it corresponds to another position on gene X (intron for example).

I hope this illustrates the issue more clearly.

Is there a way I could upload the primers bed file and bigwig files that I have for the publicly available datasets on perhaps another genome browser to visualise them together?

Answer 1

[Istvan Albert]

The essential thing to remember here is that your problem is NOT related to NCBI blast or IGV;

Your problem has to do with the discrepancy in the annotation data that is visualized in each of these tools.

Upon investigation, it seems that the two data sources differ in what they ARHGEF10L

NCBI RefSeq designates it as the region from 17,513,435 to 17,697,875

but

Ensembl designates a shorter region that spans 17,539,698 to 17,697,874 some 26K bases shorter. The upstream region is designated as lncRNA

IGV displays the Ensembl build of the gene.

You can see both genomic features on the NCBI genome browser at this link