BED alignments not positioned properly in IGV

Question

[ad]

Hi, I have a problem in IGV. When I load a BED file the coordinates do not seem to be aligning properly with ether the sequence loaded or the ruler at the top of the window.

First off for example if the start site is 1 IGV seems to treat it as if the alignment starts at the end of the 1st nucleotide. The blue bar covers the 2nd nucleotide but not the first. I was under the impression that a coordinate position means that the nucleotide start is included in the alignment.

Secondly the blue bar seem to only appear in the vicinity of the position in the BED file not the exact position as going by the ruler. This seems to be most pronounced for alignments further from the start of the chromosome. On the other hand the alignment appears to be aligned properly with the sequence itself, aside from the start issue described earlier. i.e., I know that the start of my alignment chr1:100-200 should be ATT and end should be TTA. And while I see on the ruler position its not right the alignment to the sequence itself starts at TT and the end is TTA.

When I hover over the blue bar corresponding to the alignment it sometimes gives me one or two sets of coordinates that do not correspond to the BED file line or the position on the ruler but is closer to the BED file line than the ruler position is.

Anybody know what could be up?

[dariober]

Attaching an IGV screenshot of what you observe and possibly a sample of the data might help us...

Answer 1

[Alex Reynolds]

If IGV renders BED elements in zero-indexed fashion, then the starting base may be one unit/base past the number in the start position. If positioning starts at zero, and you have a BED element defined by the interval [1, 2), then this is actually the second base along the axis, with the first base starting at position 0.

[ad]

Isn't the convention in BED to have base 1 be the start? I've never heard of a position 0 in a nucleotide sequence.

[Alex Reynolds]

I think if you put a zero-start position into the UCSC genome browser, it will add one for you, since it renders coordinates with 1-based indexing.

Answer 2

[Trof]

IGV does use zero-based index by default. https://software.broadinstitute.org/software/igv/BED It can be modified in track line, by coords parameter in track line (also linked there) but I can't make it work. Not sure if it works for BED format at all.